Transcript for:
Highlights from the Hackathon Presentations

they will change in a second so hey everyone so today we are in high quality 2.0 we had many project submissions and from there we were only able to select a few they were really good project submissions and it was really tough difficult for us to select the top few projects so today we are here we are with our judges so first of all i would like to give a brief introduction of some of our judges the first one is saurav thakur so he is a csc graduate from sitkara university and is an active open source g-soccer mentor he has a really good number of contributions in open source next hacker patel he is a mechanical engineering graduate from nitro and is currently the chief of blockchain product at meta decrypt then mansi pariaar parasha she is a cse graduate from vit chennai and has really good experience in data instrument and also she has intensive experiences as data science at many good companies then we have aman kumar divan he was also our mentor and judge in high question 1.0 and currently he is working at general electric as an additional engineer he is an experienced iot developer and uh personally saying he has helped us a lot really in organizing high quality then we have vinit and he is from vit chennai and he is a leader web developer and rescue analyst at sterling next up we have mohit parikh he is the ceo who is founder and ceo of anakin which is a y combinator season 21 startup then we have abhishek upon you he is a csc graduate from triple h opal he is he is a software engineering and he is also a software engineer at gp as well as volunteer at echo home foundations so basically to start off we'll be calling our the calling the top 15 participants and we'll be giving everyone 15 minutes time in order to display their projects five minutes sorry you know we'll be giving the five minutes time in order to display their projects so we will be getting them one by one uh i had a question uh what were the things what were the three things that were told we should base our scores on because i can see that there is only one column and well let us originality creativity and technique i am writing a chat so our first project sulpha is going to join us soon okay we have our first project stealth by pusha pusha you can present your screen to start your demonstration okay so uh my project name is stealth yeah so it's a basically our ultimate tool for the all the youngster that uh we faced in while watching the video youtube video so uh my it's my app which web app which used to just to remove all the hassles that you faced in our days so my problem statement is like uh when you blindside hear me he or she can waste his or her time while watching the video uh if does the video is productive or not so so the my main is idea to just remove the hassle hassle and define the video of the relevant content so that's why i created this uh website where you can it's a solution uh i have used nlp where when you just give the link of the youtube you will get automatically fetch the data from the link and you will get your keynotes summary quiz and you can even search the faces from the keyword phases from the youtube video so so this is my demonstration you can see it i just i just so find any video i just paste the link of the video as you can see that there are many for option you can see if you want a key note if you want to search any phrasing you can just write it over here or you can speak also this is the get word how many times this guy has said you can say also when did he said free revise the video if we have exam we can use this type of things and same as well this goes to the zoom link also zoom transcript and from the legal documents also you can just give the pdf and you will get the same all this fourteen things okay so if you want a live demo i can show you that also i just as i click a youtube video of edit and i just do the get videos as i write the keynote of the session it's it is inserting so it will go to get to me at that point when here this video says inserting and if you want to search any phrases you can search it if you want any summary of the video it will directly give you the summary of the video and if you want to generate a quiz you can generate the quiz from this video and it will also tell the analysis of sentiment uh whether the video is positive or negative it's just from minus negative to plus you know plus positive so this uh find actually from the analysis from the comments of the user of the youtube video okay so this was the uh youtube one and same as it you can do in the zoom link also so this was my project thank you thank you push any questions uh hi coach so i have a question so i'm assuming you're using a voice from text voice to text conversion right and for summarization you're using an lp model is it yes yes yes okay can you give me a few specifics of the nlp model that you're using uh nlp uh i'm using so as to uh it just uh so for a back end i'm also using the azure function also azure cloud from there uh it just fetched the uh data of the video and from this specific keywords are there so it gets the summary or the things that the user wants uh i think the question was that how uh like i don't think you touched the topic like how are you processing the video and getting the data out of it yeah just first first of all it's a youtube api this is fetching the data of your youtube video and this and there is a second uh api also as that's the ambiguous api from where uh and there is also semantic ap also where you choose to get the semantic uh analysis of the minus negative to positive negative is two to three apis and uh and use also use the azure function over there for the nlp processing right thank you okay okay next we have one's going greetings everyone am i audible yes yes you are so basically this is a website where students can learn with their finance so first of all let me give you a demo so here we can create an account or log into the 16 account so now i only have installer so this way i can sign in and and and there are many courses like we can add new courses from the database and currently i am i've added two courses these two person we can click on learn now button so here's the course like and the best part of this is that when a student watches the course like if there's a live video course so there's a feature of the live chat on youtube but when a course is recorded there's no feature for the student to chat with each other so here i have developed an application where students can chat with each other on the recorded courses also so like here's a first user and let me open another instance so this is basically an other user now here we can see that both users can chat with each other while watching the course so now they can chat with each other by watching and the user while clicking the course can also take the notes using the mongodb as you can see we have added a node and we can view the nodes from the nodes here our node and we can also add a new node from this page also and this way so we can add a new node and similarly a user can interact with other users while watching the course and the next part is that there is an image processor for the instructors so whenever an instructor gives a project to a student so basically what is what was happening in the online class is that one student copies an assignment from the other student so due to this there's a lot of means cheating among the students so here is an image processor in which i have used a deep learning ai using then using the deep eye api so we can check if two images are similar or not like i have circuit two images and they are similar so it will connect with the deep learning ai and check if images are simple so the images are similar so it has given a score of zero which means they make these are similar and it will have the instructors determine if they have a lot of assignments so we can automate this process and determine if the students are cheating by taking assignments from other students because manually checking each assignment will take a lot of time for the instructor yes i have one question so when you represented english processor right so you said that i mean the instructor can compare the images right so there is not always that we can export so let's see if i am submitting this people talk maybe some periodic issues yes actually in the in this prototype i have not implemented the pdf to image conversion but in the means next version of it like it's extended version we can create a pdf two image converter api also and then use these images to compare each other no so i think you you misunderstood me so my question was actually we should look into the plagiarism checking them so in his comparison is just one thing so that can give you an idea of like similarity between two uh two assignments but uh don't you think like we should have the plagiarism checker let's say if you have given me some assignment and i directly copied it from some google visual although that image will be unique in your database because you won't be able to find that because nobody has copied that from one right but actually i have copied from the google resource so don't you think like we should have the plagiarism report rather than image process yes we can have that also the platform checker yeah we can also edit but actually i was having less time so i could not implement yeah we can implement it any more questions uh yeah i have a question uh punch good project by the way uh i want you to know what uh stack you are using for the website and like i could i can show i can see a video on the course is that just a uh just an embedded youtube video or are you like uploading videos to your website or what's the what's the technicalities actually i'm using the mongodb on the back end dynamically i'm loading the courses but as in the demo i have embedded the youtube playlist but i am like let me show you the mongodb also yeah okay time is up please the last thing i want to know is have you implemented the ui by yourself or like have you used some pre-existing websites i've implemented [Music] okay yeah if we have time i have one more question yeah so um right so you are screening the youtube videos i just want to ask like you also implemented something for stories of video and streaming the video on your platform itself yeah we can implement yes we can implement the amazon aws yeah like s3 storage to store the videos if we don't want them to upload on youtube but we can also use the youtube servers to upload them and keep their privacy as unlisted because that would be more efficient way but we can also use the our custom storage like aws or buckets storage buckets thank you ones next we have underwater uh why is my screen visible oh yes let me see the first slides too so this is this is the demo of my app kodi so code ai it's it's a web app that is a peer program it was built because we wanted to help developers and students to be more predictive and to educate them so so it contains set of tools that can help you to write code faster and better it can help you to write the sql queries javascript typescript code regex it can also convert your code to different language it can help you to understand your code and explain it to you it can also calculate the time complexity of a code and it can also prepare you for interview by generating integrations so how we build it we build it using next year is typescript open ai api and delivering css for styling and virtual edge functions for when we used virtualized functions for server what you say i cannot explain it but so here's this first infrastructure of the app so so it's i let me demo you the app so here is my app at domain code.tfs.live so let's get started so here are some of the tools like mysql query generated javascript generator code explainer methods so let's try first mysql queries and twitter one so i have set up questions pre-written so let me at first the query to select all users with id one copy it gave us the sql com command for that so select start from users where i discussed one we can also copy to the click button so we show you other tools like javascript and typescript generator it is currently in beta grab and present create a type clip function to fetch pokemon api so here is the typescript code to fetch uh pokemon api excuse me so uh can you change this to like in javascript just write something like write a polyfill for uh promise.all this query so write a polyfill for promise.all i just want to say like you know how it's behaving write a polyfill for promise.org only fill a qualified promise is the output uh yeah i mean it's actually not writing the polyfill which it has given us the promise knowledge so yeah i was just checking how does the open here here it behaves okay it's fine please uh the other one is a code explainer so currently support javascript and python uh let me grab and code like this is a this is a python code to generate fibonacci series the code is the function that prints free markers equals up to a certain number it works in this also next tool like interviewing generator and i could give give a topic and it will generate interview questions upon that like react.js summit so it gave us set of reactions there's some bugs also here's the main is a great feature regex generator so here we have to uh put a uh string that that will be on our search like search box like like we can generate uh regex to authenticate data string contains these three unique values so here's the legs for that time right so another one is a time complicated calculator so let me try to code for this first normal print hello world the python one it will be all not n so the time constant response is one naught and then n is the length of the string and let me i have a code that that this is series it is a recursive function the anti anti-time velocity is go not to the power n so let's try this so it gives it give us the correct result so let's try the last one it's the prominent feature like it can convert the code from one to another let's try first python to javascript and here's a simple python code so console.log hello world let's try that uh fibonacci series one sorry to interrupt pandora please be fast you are running out of time okay so here's the fibonacci one so it generated a python code but there is some regex error it is not formatted correctly in python but it's great so here was my project so i think this is kind of a basic version of what we have for co-pilot tab 9 etc yeah yeah so many uh code ai pair programmer for us so i just have two three questions right so in terms of usability right so it's a your your core ai programmer is actually hosted in a separate website right so but i think like it would it could be great like if you could build a vs codex just as this yeah yeah yeah it's that next feature i was thinking about yeah so because no development starts from there right so you never write it on on browser you write it on editor so that was one the second thing was like let's say now uh the whole set of these features are dependent on open air so they have the database they have trained more model but like helping open your in terms of like data collection or training of data uh in my knowledge i think uh the open ai also collects your response like your api request and it raised its model upon that yeah that is correct but now how you are ensuring that the same user is getting that that sort of you know like a similar output all the time right so in copilot what happens like you are always logged in to your github account now open here understands basically copay like to understand that okay this is under log who is writing this code so let's give him personalized results yeah like in future we can implement user authentication also like it and it can store the related data of the user yeah so have you thought of that like building that yeah yeah i i thought of that i i was like building it on progress uh using super base for that but i did not implement it because uh it will be not accessible if yeah like it will be not accessible for all if we add user authentication it it would be greater if we do not have an exist and user authentication it would be anonymous but for improving we can add a user authentication also got it uh one last question so for open ei do we have any rate limiting condition or it's always free for everyone no no currently uh the currently i'm using code dvd version two that is one model currently it is in trial so currently we have unlimited request but uh generally open eye has a rate limiting uh for that for usage of api thank you thank you so much thank you yeah uh i have one last question okay sorry no way next we have project imagined by demanding team please yeah just a minute am i audible hello yeah uh can you just repeat the project name i think it's imaging because it is imagine 3d i think uh between its misspelt imagine 3d okay okay yeah okay i'll just present my screen go ahead so is my screen visible yep yes yes it is so uh first of all greetings to all the judges uh and thank you for giving us the opportunity to present our solution uh in front of you so i'll just like uh to give a brief intro of my team so my name is ridhiman and i'm from srm institute of science and technology and uh i'm from team hello world 404 and accompanying me or three of my friends uh chief malar dixit in january and naman sharma and we have built a solution in the open innovation category and i will just give you a brief overview of the problem statement so uh our problem statement is that whenever we are as students or as budding professionals whenever we are trying to understand important concepts in from textbooks we often face a lot of difficulty in understanding them because of the lack of visualization or proper denotation of various components that are involved in a particular diagram or maybe a particular theory or maybe a particular concept so what we have come up is a machine learning and ar based solution that will help you to generate a 3d model of whichever concept theory or object that you want to study about we have used the concept of ocr and we have also used the ar concept in order to generate the 3d models and this is how we have come up with imagine 3d so the main functionality of our app includes to scan the image and extract the text and get the 3d model for the desired object so we have also the facility to create and store notes and take short quizzes to improve understanding of our subjects so as for the tech stack we've used blender and 3gs for doing the augmented reality based stuff we have used tesseract in order to do the uh ocr analysis uh we have used react native to our front end application and we have used node.js express and mongodb for our backend and also html js and css for our webv so here's a basic workgroup which i have just described in brief and what we basically do is after login and sign up without the basic operations we go to the scan and learn section where we scan the image and after uh uploading the image the ocr model will detect the word the keyword and we'll send it to the uh and the app will send the request to the api and after that the api will send the response url to the application that will redirect to the webview which will have a 3d canvas having the model in place along with orbital controls and labels there is a note taking facility and also in our app we have a quiz taking facility that's present so here is a basic ui of our application so as you can see there were scan notes and quizzes section and we have a recent searches section as well and these are just some of the screen grabs out there and now i'll present to show the video times justice so can you see my screen yeah all right so here's a demonstration of a scan and generation operation so you can see that our camera has gone live we are trying to detect the word heart with the camera we have also given the crop and the resize facilities to help the user to pinpoint the exact work that he wants to search for after which we go and there was just a small layer of the list yeah so that's how webview our app will render so we can see that in 3d canvas the 3d heart has rendered where you can control it using orbital controls and you can even zoom in and zoom out over there you can interact with the model by pinching in and pinching out and you can even see some facts that have been generated over there after double clicking and here you can just write short notes if you want and by giving a title and content yeah this video is just going to end in 30 seconds or so so after that you'll be ready for the questions okay so that was just a short quiz that you can take and there you can see that the object that we are the note that we just created has been saved out there so this was a short demonstration of the application that we've made and hope that with this we have come to the end of our demonstration thank you very much we have one question so yes [Music] and understanding it's just for better analysis of a particular subject uh so that's why we've added so we are planning to add more uh domains and quizzes across more domains so as to give a better overview of whatever the person has learned as such so that's the purpose of the quiz that we have given a second question i have was like you know you initially you demonstrated that there are recent searches which is which has been shown there and also the nodes are getting stored somewhere right so i want to understand the architecture like you know once you scan the image you crop it uh yeah like that is done through the android library but after that like do you upload it to your server and send your cd yeah to the model or it's just like you are sending the whole basic no no no this is how you're doing it we are not sending the base64 encoded image we are sending the image that has been uh clicked in the same mind type as it's been generated that is jpg and we are sending it to server after it is sent to our server we are using multiple to temporarily store it in a location in our server it's an uploads folder if you see our github repository in the master branch the architecture as per the architecture we are storing in the uploads folder after we are storing it in the uploads folder we are extracting it using we are extracting the path using the fs library and then we are sending the png image or the jpg image whichever has been captured uh to the uh uh tesseract.js library uh where we are using the recognize function tesseract.recognize and after that the tesseract library uh does its asynchronous process and after which we get our uh text whatever we have received after we get the text we use fs.unlink in order to delete the image that we have stored on our uh uploads folder so as to keep proficiency and also to help the server occupy less space and uh thus after we get the text generate the relevant url that we have that we have we have a set of urls that are there in the public folder and then after comparing which organ has been there we send the relevant url back to the front end so after front end receives that url we concatenate a search query a javascript search query into that url where we get the username username equals whatever username is there in the local store after that we send a request to the webview where get request is there so there uh which whatever we are rendering we are doing a temp uh server a server side rendering only so in the server side we are using uh window.location.search we are using the search parents in order to extract the username from the search query and after we extract the username whatever notes that person has written in correspondence to the diagram we use uh another postdoc that we have in our node.js so in mongoose we are using the user and username query that we are sending in the post request and we are also sending the title and content fields along with the data that has already been sent in the post request and are making a note object that note object has been pushed into a notes array in the mongodb database and in that note saying uh there is a schema type note where the title id and content is stored so and after uh this entire process is done we get a response 200 response from the server if we get a 200 response we use the document.queryselector object to update the text area to node safe successfully after that if a person wants to view the save nodes we are using a get query get request to our server slash notes so from there we are deriving all the necessary nodes array that is there in our mongodb with respect to that user so that's how the rendering and enter architecture of the node is coming it only yes thank you so much for detailed information uh i got an idea of like overall architecture just two questions right so one is like you know i might be wrong but why are we sending the whole png image so can we use cdn url in that library that you are using and that cdn url can fetch from the image server and then do the processing and second question is like i i didn't understand like you know you mentioned through the whole architecture everywhere you were talking about like username as uh query param but how about the authentication so uh have you implemented any sort of on the authentication or it's just like if i know your username i can switch all the data that you understood uh now as of the prototype we have not implemented any authentication but surely when we are going to do a release version we can have a authentication over there where we are going to pass a uh where we are going to pass a hashed username onto the particular search query so in that manner you will not be able to actually enter the user name directly in the search query as we will be using the bcrypt dot compare function at the back end in order to resolve the username properly so just for demonstration purposes we have used the username query directly over here but uh surely we can handle this at the back end and even during registration uh we have used the bcrypt library in order to encrypt all the passwords that template so that's uh going to be a feature that will implement surely and as of the first question uh as of cdn the approach yes we did do a little bit of research about it but due to lack of time we couldn't implement it though uh one thing to mention uh we had tried the base64 encoding uh part of the same but uh malta actually uh which is our file storage uh software that we're using it does not allow a large size uh file to uh to you know get stored in the uh uploads folder because we are using the dots a single attribute of the same and also uh we try to use a large number of uh we try to uh take a large number of pictures of the same uh basically uh alphabet or the word but uh we were still getting the same error but surely we find the work around it good yeah i think that was the reason i asked these questions i want to understand you know what is the exact stream uh i have a question which is like uh besides the technical stuff that we are doing here how are you like from where are you getting the ar models that you are showing to the user so i don't know how do you yeah yeah the second part of your question actually so i also wanted to ask let's say how will we like add more models to the library so how does it all happen okay okay so this is how it works actually uh as of now we have uh got some uh we have we have uh got uh we have worked on some previous projects so uh in in blender actually so we were able to generate the 3d models using blender it's a software so uh what happens is uh blender generates a dot glt file uh so the dot glt file is basically what we use uh in a glt uploader that is there in uh 3gs uh the server side uh that we are using that will be helpful if we just talk about the like the overall things uh like from where are we getting the models and how will we like expand the library if we want more models the overall picture the scalability yeah okay scalable all right so as of the scalability aspect uh in order to get the models we surely need to have a dedicated team to work across various domains in order to have quite a bit of models ready now acquiring models uh and building those models will definitely take a little bit of time but as of scalability the thing is that uh we can easily use the uh file formats that that is converted uh over there like we can convert those files into a gldf format and after that uh we can use load balancers on our servers to uh deploy such models that are large scale so uh that's how we are actually planning to do it uh one really quick follow-up uh rhythm again thanks for the detailed explanation so a follow-up on saurop's question you have a database where the 3d models are stored with your query as a key of sorts so you're not actually generating the 3d model based on the ocr recognized text right no we are we have pre-generated models we have pre-generated models uh but the model is basically uh just the 3d object the animation the uh orbit controls and everything we are adding at the back end as per the user so ocr will detect that and will generate the relevant query to us at the back end where we can use any questions that will that can be done in a custom manner got it got it thank you we have project soleplay by team blitz and we're fitting to present hello everyone hello hello everyone uh i am khushi team member of blitz so we will be presenting our idea together my teammate pritish will be presenting presenting our app so starting with our app uh we have made a soul pick crypto wallet it's not just a crypto wallet wallet it's all in one crypto wallet for all your need we had we have made it safe and easy for customers to store buy send and receive money and crypto so starting with what the idea behind our soleplate the idea behind solfe is to bring the power of solana blockchain and wallets to as many people as possible the service can be can also be embedded into any website with minimal code it allows that website to create and track solo solana virtual accounts starting with the technologies we have used in uh our app we have used javascript next js web 3 and blockchain uh going through the uh facilities we provide to our app uh it's a home for all your nfts you can buy mint and sell your nfps with just few clicks you can earn cashbacks swap with ease access daps pay 2 merchants and get get cash with ease we also promise secure crypto wallet by generating password and keys on your device so only you have the access to your accounts and data we this our service is built using react.js for the front end and and next js for the framework now we will show our app demonstration uh starting with creating our wallet in our app then we click on get started here we create a wallet here a phrase is generated that is totally random so that no one can ever uh copy or hack it that is provided for your security uh so clicking on create wallet now my teammate british will proceed just give us two seconds please so this is our wallet solby uh first of all uh for testing i have added airdrop version so that so let's get started so now as you can see my balance is around one sole features again we can receive money through the qr code or our wallet address we can send money and we can also it's a complete wallet system with uh nf nfts uh we can also maint buy and uh check on our nfts that that will be buying soon uh we can receive money through our qr code section we can also track the payments let's see so as you can see transaction was successful which was you know that airdrop we got before starting about sending money we can write up our value and also put up a wallet address or we can scan the qr code so this is a link to the pair we call it and we can send him the money so that's all about our application thank you any questions you have uh yeah i have one question so so lana pay if i understand correctly that's kind of a standard protocol protocol for this uh decentralized payments right yeah so you you use that open source library that helps us to build this uh particular wallet thing right yeah yeah sure so can you kind of briefly explain you know your work in terms of like what you have done exactly so so lana pay already provide infrastructure to build balance so uh apart from that like you know what you have done uh uh for that we have added some features which we are lacking in current crypto wallets like uh we can receive so i have tokens uh from this only and also collect nfts or buy nfts so that was our part since we wanted our crypto wallet to be all in one crypto wallet uh so that there's no need to switch to some other applications or projects to do other things and then for the ui uh like have you used some template or you built it from scratch we build it from scratch okay but i could find some lauren episode uh uh actually that was part of my production [Music] questions from my side i went through a general like when you talked about the value addition you had you said that your uh your wallet allows to swap tokens right yeah yeah but um even phantom even phantom wallet the extension for which anything they have the features of storing anything so what exactly is your valuation figured out before this and if i and uh you an address comes up right like but i registered just now in your website and then i got an address but i can't seem to copy it and i want to check that address on soul scan about its status and how much salon i'm getting right uh yeah oh yeah can you repeat your question like okay see when i when i entered uh when i created a wallet i got an address so when if i go and search this address on solana scan sol scan then i should be able to see the status of the account with the account balances and the transactions and the whole history right like what happened it scan polyscan right yeah yeah so a i'm not being able to copy that address and b if i were to be able to copy that address will i see that i was dropped once all in my account so we have added a transaction history when you click on transaction history you might be seeing the transaction of that but see when you air drop any solana into my wallet that's a transaction right basically so it it should take time to but because when you when you tell me to see the transaction history on your wallet there could be a chance that it's centralized right in your github code base also i don't see any rust and rust is the primary your language for smart contracts it's okay uh could you just give me access to the you know address because when i copied and you know the doctor dot is coming i'm not getting [Music] like for any wallet any wallet whatsoever there is a option of networks right because other than the mainnet there are test nets involved test network which you have while developing but when i enter a wallet i can't find the same i can't see the options of different networks right that should be provided as a basic feature of any wallet you can take metamask for an example metamask will have what you can add whatever network you want but that's but that's the basic like right by or robstern or govan or worldly test networks right solana also has the same and yeah it's not provided in your application so you should also work on that okay uh i have a question uh and like in the starting of the presentation you told that uh your wallet is secure uh can you explain like how are you doing that because i can see first of all that you are using web app it is a web app and like you also told that you are storing the credentials in like the user device like how are you doing that uh uh first of all solana blockchain itself is a delicate uh thing the data storage is happening in from dedicated validators to archivers so when any a small transaction or some uh archive is passed so so it goes to it goes to our network then this for our data privacy we have proof of history solana blockchain actually named me mainly solana blockchain has implemented the proof of history which is which is time keeping it keeping technology kind of so pratish uh i think you are talking about the transaction security i'm talking about the user credentials here like where how are you making sure about the security of that that i told first the data story dated from validators to archivers only so uh whenever a small section of the state is archived by us uh we will be only asked about uh to confirm and uh storing the data so well i i didn't get it actually because i understood sarah's question like that is a serious issue in all the wallets in which you store their private keys because i have a private key generator based on which i get a wallet right and based on which which is the identification for my account so where are you how are you ensuring that this private keys aren't elsewhere because i sort of had mentioned that you checked your code base and it's uh js yeah we should be working on that um it's the start of this so we were still working on these parts of our phrases oh okay okay thanks for answering that and you have a lot to work on here because there are tons of projects with saltwater i mean it has you can just go to get a bunch and check all the projects and software there are tons available right now so you have to make it really unique if you want to go forward with this project else you have to find another use case any more questions okay thank you next we have project yvonne by team errors uh hello sir sorry uh it is for team errors right sir we are teammates sorry the project name is fearless okay sir thank you soon so this is our project it's it is name is phonics so basically what it did is we basically provide the solution for developing ev infrastructure so what are the basic problems that that make us focus on this problem so first of thing is that that india's ev business faces a lack of infrastructure for evie and next thing is that india has a widespread charging infra like the demand for ev vehicles is growing very tremendously because the fossil fuels are depleting a very fast rate and people are also struggling to find the closest charging station that why they like they don't want to buy ev car or something because there is no infrastructure present so like what are the constraint when we want to provide our solution so what should be their constraint the user should have to travel minimum distance first constraint second thing the cost of installing ev charging machinery should be minimized i mean that the with minimum number of ev charging machines we should uh satisfy the most demand and the how we can do that if we better predict like what will be the future demand in that particular region using that we can like optimize our course so what we are presented so here is our website so [Music] how we have proceeded in it ah let us reload uh it's taking time actually ah yeah so here is our website so how we have done this so first [Music] we have calculated that what our target area so first we have do some data analysis and find out which are the states that have the most emerging demand so here you can see that the states which qualify for most emerging demand are uttar pradesh and maharashtra so it is the vehicle counts in that region after that we have targeted uttar pradesh so going in that we have forecasted that how in the upcoming year the demand will change okay so we have data for 2019-20 and we have forecasted the demand for 2023 and 2024 so here you can see that the demand has grown tremendously because uh the fuel prices are going to like increase very fast because there is a depletion of that so based on this uh this is our forecast we have used xgboos for forecasting it and here is how we have clustered a particular region for assume that i have selected a 60 crore 60 field and in that i have like every supply point is a petrol pump and it will cover a region around it so we have used k means clustering for doing this and we have made it 100 supply points like there will be 100 like 100 points where we can install our evs okay so for example this is yellow point so this is a supply point it will meet the requirement and demand of all this region and moving on like our now next problem is like how we can provide the optimal number of evs so based on the demand we have predicted it we have assumed that one ev charging point can satisfy a thousand megawatt of charging in a region okay so here is our map that will tell everything okay so for example let's consider this area so this is the area that have 100 like those 100 supplier point so this cluster involved 36 petrol pumps this 31 this 23 this is 10 so let's go inside some uh this is 36 points so this region includes six points and at each point you can see that this area require 18 uh charging like 18 ev machines at this petrol pump in order to meet the requirement around this area similarly this needs 16 this needs 17 and similarly we have done this so so after that we have also provided a heat map that will tell us how the demand is like how the demand is distributed so the more red region included here is more demand so we can like also predict like compare them like if our prediction is right or wrong okay so after that uh we have also included uh like we have also included a feature for the user so basically what it does is it for if you enter the coordinates yes sir you're running out of time please be fast okay okay thank you so in this uh if a user enter the coordinates of its location then we have we will tell him that the shortest distance like shortest supply point where you can charge is at this kilometer and the coordinate so you can easily use google map to reach there and you can like you can use this so this is all we have we can also optimize it according to user needs thank you sir that's all from my site thank you any questions uh i have a very quick question so can you tell me about your data like where did you get this data for indian cities yes ma'am that is a major problem uh actually india has no such infrastructure for ev so we don't have such data basically we have tried to like optimize if government can provide us data we can use that to similarly we can use the data here so there is no problem in that like we can work on that but we have used uh some data set maybe it was of germany and we have used that data to implement it here so basically we are so by focusing how we can implement this okay yes got it nice good job thank you any more questions okay thank you for your presentation thank you next we have ergonomics and emotional wellness by team dns and we have recon to present deepen yeah am i audible yeah yeah you are yeah is my screen yes it is yeah so the project we have chosen for this hackathon is ergonomics and emotional analysis so basically chosen the healthcare the problem statement that we have chosen here is that so people working in corporate environment so basically in cubicles or maybe in a work from home environment uh you usually forget about their body posture and incorrect body posture for a very long time can lead to severe conditions like musculoskeletal disorders and poor circulation and impaired lung functions and this project was made with having the you know the lockdown and welcome from environment in having those environment and the solution we have come up with is that so we have come up with an application uh with underlying uh two major algorithms so basically it works as an assessment tool to you know i identify a risky posture so we have we have implemented two industry standard algorithms called ruler and riba so ruler is rapid upper limb assessment and rebar is rapid antibody assessment and the unique selling point is that so we have used deep learning models and we also use the media pipe to get get the poster plants and most of the economic related software's present in the market right now are basically static static image and video based softwares but here we are doing it dynamically with with our webcam based data so this is how the architecture would look like so these cameras can be considered as the uh the webcams of people who are working in cubicles and the these from here the posture points would be collected and they will be sent to the server and after calculating calculating it a a notification would be sent to their mobile whether they have correct first round for example correct post or not for example if you are leaning two right or two left you'll get a warning or if you're feeling a bit drowsy uh you'll just get a wanting to maybe take a break or something so that's how it looked like and this is how the flow chart of the working looks like and uh so these are the technologies python opencv media pipe and tensorflow so yeah i'll go ahead with the demo right now yeah so basically you have a login page so so after after logging in so face recognition data does not exist so you need to register your face so you have to go here and you can you can just give whatever id you want so basically if if you're going to implement it in on a company a wide company based range so the company would be giving an id i id too but for the sake of demo we can just to choose a random id here so when i am starting a phase capture so now only your face will be detected and so these images would be used to train the model for face recognition so total of 80 images will be taken yeah so the phases has been trained so next you can go to the detection part yeah so basically this project contains three parts the first one is the ergonomic analysis the second part is emotional analysis and the third part is browsing detection so the economic analysis basically consists of the uh checking your neck inclination and torso inflammation so if you are going to uh tilt your neck to to left or return it to right it could be detected and a warning would be sent to the user through telegram so just yeah so a warning is supposed to come here so that's how it basically works and here and the middle part is basically uh detecting your emotions so with respect to emotion you would get you would not get any warnings but uh you know it could be used for long time analysis of of your emotional state through to the working capacity and here here you would get a drowsiness detection and you get a live graph you know whether you're feeling drowsy or not and we have also incorporated face tracking it would take a bit of time yeah so why we here we are using face tracking so when you are going to use it in a work from home environment uh case tracking is not necessary but in case of a cubicle based environment definitely two or three people there is a possibility of two or three people coinciding with your webcam so in that case you can use face tracking to you know accurately get the results so yeah so after uh you know after running this software a report could be generated so this is basically just a rudimentary uh you know proof of concept like report so like this a report could be generated with respect to days data or months data so it basically consists of you know uh neck inclination torso inclination process score and all the emotional emotional score so these uh the inclinations and the division score can be used for your personal development and this uh emotion based scores can be used by the human resources or your managers to assess uh about your emotional state during the pro during the you know the tenure of the project and maybe they can give you counseling um it can be used for uh you know that purposes so yeah that's pretty much uh is our project so yeah i'm open for questions now okay any questions i have a question so deep in a really good project let me start with that so my question is more related to the use case from what you have shown you basically want the employees to have their webcams amazon uh 24 7 when they're in office right yeah yeah yeah uh yeah we get we get that a lot so basically what we what we are doing is edge computing here so uh your webcam uh the frames from your webcam will be processed to get the uh the points of your shoulders and it it will all run within your computer itself so the data which would be going to the server would be only the coordinates of the coordinates of your shoulder and your trees and moreover uh there is another there is some types of webcams which which solely works on infrared uh you know a different types of image processing based cameras are available so we have only the webcam which which is given with us the computer so that's why we are using so in case of some special types of cameras uh only the outline outline of the person could be captured so in that case the privacy feature could be the privacy could be preserved so that's the solution we came up with so we can't implement it right now but it's uh i would say that it's a good solution right right precisely my question was how would you protect the privacy so you seem to have that covered thank you any other questions any questions magnitude decentralized petitioner by team binary brains and are you showing present ayush you can share your screen ayush okay seems like ours is not raising we will move to the next project we have project e1 by team bit by defeat and it will be presented by delhi do we university and our theme is uh ev charging station which is an open version uh hashimoto is also in my team he's a web dev and our team name is bit by bit so we know that like ev is the future of auto mobile industry so does tv charging stations are uh but after going to a survey we went we were about to know that 30 percent of euros forget to charge the dv and around around 40 percent don't want to own any because lack of av charging infrastructure so we are trying to solve this thing by introducing avion uh where we are redefining ev charging station so currently there are a lot of problems but these are the three major problems what we saw in the ev infrastructure this is searching long waiting time and very limited public places to install the evs charging stations so how are you going to solve it so our flagship feature is home charging stations so what is home charging stations here any ev owner which who own an ev have a home charger so using our application he can register his home charger and he can start lending that home charger to people who actually requires ev and then he can uh indirectly start his own charging station with a zero zero investment and start earning from lending his charger so this is our flagship feature which we'll be providing the next thing is we are trying to connect this ola ether bounce chargers location in one application if you see that there is uh ether ola they are trying to create their own infrastructure or application but we are trying to collaborate all these things in one application so that the user need not go to different application to search for different charges the third thing is we are trying to collaborate with the citoria mall and universities where they can easily install fast charging station from our application using small process which will be directly linked to the companies who will be providing fast charges so here the cafeteria mall our university people can easily install the fast charging in their premises so these three are the major idea of our application apart from that we also have a feature uh which is known as av algorithm so what this does is let's say your ev has a low battery percentage so what are we gonna do just run our algorithm based on the traffic factors destinations and write a starting step it will calculate what will be the nearby charging station you have to visit up to what percentage you have to charge your av so that you can reach your destination in a limited estimated time so this is kind of proud like a unique feature which you are adding in our application next thing uh there is also hardware we are trying to build actually uh which can automate the home charging station if you don't have this hardware you have to manually go scan and provide the charging facilities but using our hardware you can automatically automate those charging stations just by keeping your charging station over there and the user will come and he will just charge and start using the ea charger uh so like now how are we different like we have the home chargers which is very unique and we are the only one who are providing this feature currently uh there is multiple charging collaboration as i said ethercola bounce and we have the b2b strategy we also have a business model over here uh there is two business model there is b2c and b2b in b2c as i said we are dealing with the home charging station where people will land their home charge and they will earn from the in which we'll be taking some cut and uh as i said we will be collaborating with my coffee cafe malls and uh universities and all will make them collaborate with the companies uh to provide the fast charges in which we can take some control so this is the two business model which we are planning to implement in our application uh like how are we different from other providers let's say ayla or ether ola bounce and tata uh the first thing is as i said home charger this is the unique and flat feature which only we have the next thing is all in one search where only a tata has and we are also implementing it next thing is easy installation uh as i said you can easily book the fast charge installation from our application low battery algorithm this is also a unique feature in our application charges live status which is almost all application givers next thing charging monitors or you can monitor your charges health and how is it working what is the state of the charge so this is our usb compared to other provider yeah so our tech stack currently just to show the prototype is the flutter as a front-end firebase as a end and arduino to show the hardware demonstration and there are a lot of use cases but these are the three major use cases of application let's say av home ev home charge and fast charging points uh people can start earning from providing their home charges and the easy installation part so these are the three major use cases from our application our future goal is to also include the battery swiping feature as we can know the bounce has introduced battery swiping feature so we are trying to uh implement this in the future because uh let's go for a demo we have a small video i'll just show it to you yeah so these are our move user application whatsapp application you can see that the person can either search for fast chargers or a home charge in the right side uh you can select home charger you can search the home charger and in this app again you can see the reviews all those things in the map you can see where to visit it and after the confirm you will just get your estimation time and after confirming it will get a request in the merchant application where the merchant will accept it or reject it you will accept it and then you will get generated xqr code and after that you will connect your hardware uh this is just for prototype it's not the actual hardware uh where you can just simply scan the qr code from the merchant application or you can just leave the qr code on the hardware and it will just scan it and it will just show the charging point and it will start you can see that uh the red light here it's started charging and after the charging is done once the charging is done the payment is generated and the user can pay using the rates of the application we are using currently and after doing a successful payment uh the bill is generated and sent to the emergent application yeah so that's it from us like we are open to [Laughter] thank you from my side nothing from myself yeah nothing from my side as well fine for your presentation oh yes sir warm greetings to one and all present here we are here team bitwiser and we are here with our project secura which is a woman security application and it's like a private security guard of a woman and through this application we are creating an initiative that focuses on women's security now i just want to take you quickly through our team our team consists of utkar sarura shivendu mishra yeshweed and myself moving to the next slide we have chosen the track woman security because we feel after a lot of tremendous advancement in technology still a woman when travels at a public place or in any public transport she does not feel safe and secure she fears while traveling in public places so to eradicate her fear and to make her feel secure and safe we are here with our application secura which is like a private guard for a woman so this app mainly focuses on problems faced by women and which becomes a great hindrance in but in becoming self-independent the app provides some great features which can make use of anywhere anytime women's safety in public places is our first and foremost priority which we are trying to solve through our application also we have integrated a simulation in which we have tried to show some defensive techniques through augmented reality arvr to train her to make herself independent so that she can fight for herself so basically through this application we are trying to secure a woman so now to tell you about our features of the application i would like to call my teammate utkar sharora thank you ssv good evening everyone my name is utkar sharoda and today on behalf of team advisor i will be going to tell you about the features of our app secure your own virtual card so firstly please open the app chimindo so firstly we have implemented the google sign in button so that the user can eventually log into the app securely further on moving towards the home screen yeah as we can see the home screen we can see the emergency call feature with the help of this feature if by chance any girl or a woman is stuck in a critical situation or needs urgent help then she can just tap on the button and the call will be sent to the trusted contact which is indeed a very important feature but imagine if the situation gets worsened and she is not even able to press the sos button then she can just shake the phone and throw it away the phone will detect that there is a problem and will actually dial the emergency contact in next 30 seconds this feature is one of our greatest talking points as this feature is nearly similar to crash detection feature which is introduced in the iphone 14 series last week on moving further if a girl is traveling alone and finds any suspicious behavior around her then she can just call any of the national helpline numbers like police ambulance women like power line numbers etc which are provided by the government now imagine if girls get proper defensive training then she can prevent mishaps so going with the idea we have implemented the real-time simulation of a girl practicing defensive techniques with the help of augmented reality so here is the demonstration of our augmented reality app firstly it will detect the plane as you can see there are small small points which are coming in the camera then it will detect the plane that there is a plane and then we will put our characters on it now as you can see the woman is trying to do some defensive techniques and man is all down then this is the simulation of our main defensive techniques and then if this feature comes in mainstream then it can create a revolutionary impact in the safety of the women now we have also tried to add a list of trusted contacts if you tap on any of the contact cards then secure will automatically call the person in case of emergency now as you can see we have added four contacts as in a scenario suppose radha is a girl and she is stuck in a totally unfamiliar neighborhood so with the help of secura she can easily track herself with the help of gps and live geolocation services and if she wants to share her location to her near and dear ones she can also easily share it with the help of the google maps now as you can see shivendu can you please go it further and you can press the back button and please move press on it again as you can see it shows the live location of the person and can you please tap on the red marker and as you can see we have redirected to the google maps and through this we can share our location to our near and dear ones at last we have also added the woman power line official website in secura so that women can have a check on the latest updates and notification from the ministry with the ease of just one click now as you can see we have implemented this web view in our app now let's have a look on the tech stack we have used in the project okay so the major tech stack we have used is kotlin and xml and google air code for the augmented reality modules the firebase authentication with the google sign in and the firebase real-time database for storing the data the ar and the unity for making the same ar core modules now my friend ssb will tell you how secure is different from other existing market competition over to you yes ssb yeah okay thank you so yes we know there are still many existing apps which are there for women's security but the ar augmented reality feature which we have added is the most differentiating feature which we are having in our app also while shaking our phone if a woman is stuck and she is not able to call or even she is not able to message her sos contacts then the phone then this is the shaking alarm which will automatically run so yeah [Music] feature of our application now also we are thinking to implement a good road bad road indication uh through ml machine learning algorithms this is our future goal uh this feature is going to tell a woman whether the street she she is going to travel is safer for her or not so this is our future goal and yeah moving on to the next slide so thank you so this was our application secure which is trying to make a woman feel safe and secure because this is a major real life problem which needs a static and urgent solutions so yeah questions thank you for the presentation judges any questions well i don't have any questions but uh i think uh like i like the whole idea and you have tried covering many parts of the year thank you and i think you should focus more on the ar part because currently the animated right and you if you want to teach you know the girls about self about self defense then that would require a bit more refining and good job with the with the sos call and the security sir i would like to tell you something about the story of the making of the app so actually we took only 12 to 13 hours due to time constraint we were only able to add the basic ar model so but we have decided in future goals that we will add some more and more models which are complex and which which can bring a realistic experience to the woman so that she can learn at our own pace the self-defensive techniques yeah but that seems like a good idea just while building the app target the simplest feature first and complete that and later go on to the more complex feature it gives you constants any more feedback or questions just a second guys uh i have to leave uh thanks for having me here and i will review the recordings and update the rest of the projects thank you for joining us all the best guys thank you sir what was the name of the last project which the one which just played so while building uh while searching for teammates uh we realized like me and shaykh singh gupta realized that we were actually there was lack of partners who were actually who will team up with us like while we were actually building the projects or something like for this for this hack audition only so we came up with an idea to build a project where we can actually find partners for hackathons or we can save for projects so we have built this website where we can actually find teammates or we can say partners for hackathon hackathons so as we can see here the code base is here in the github so this is the website uh so at first i will just sign in here yes as it is logged in so here we can see is the profile is here as we can see that their profile is here and in the dashboard we can see that name and some people who has actually searched here and they has came here like for example if i want to find a teammate for my for a hackathon we can write here like uh for example hack kodi shop and my mail id then html css like what are the skills we want in a partner actually javascript and this is optional but we can give you like i'm from third year so as i can say it's as we refresh we can see that uh it has been updated so from this we can actually have conversation with uh people also like for example if i want to have conversation with this person like my other account so we can go here and then we can actually have conversation with that person like for example okay like we can have conversation with that person about the hackathon and then in this website we can actually edit the all our skills and stuffs like our banner or we can actually edit our infos like what are the achievements we have done here so we can actually do this thing then we can actually find some hackathons that has been presented here like we have put up here using our api and then this chat portion and this uh faq where we can actually talk about how it is done so this is the simple website that we have built here uh with my friends that is developer and me as a full stack developer and tamaldas as a front-end developer so this is our website and here we can see that how the workflow is done like since we got much less time so [Music] we didn't uh got time to prepare uh canva so this is our website so a simple website where we can actually find is this different from the current uh portfolio system of let's say therefore you know yeah for you for example okay so when we go to date for you like for example if you are searching for day for you we can actually get only the hackathons that are there but we cannot find we cannot find any partner section like for example if i want to search for partners who are actually who want to build together who want to be this is like a facebook or i can say linkedin for hackathon people hackathon enthusiastic people i can say so this is different like here we can see that there is some hackathons open but we cannot find people who are actually interested in hackathons so this actually helps this our website actually helps to find people who are actually find teammate for a hackathon say eat global they do that thing they have a dashboard if you're if you are searching for a teammate or a team then you can just enter your details your specialties and then the dashboard gets empty of course we can uh do that thing like dashboard where people can actually join but we needed to get more features for example we can actually put up this kind of stuffs and we can actually talk about only those portions which are actually important for our projects or i can say what for only those portions related to hackathons that's the reason we made this website that can be actually talk about only hackathons only [Music] and uh for further forward technically i think it's at a sound stage right now but i would suggest you look for the target audience for your product because generally like people who are very active in hackathons they have a particular team in which they always go with right a number of number of people so your target audience would be the newcomers right so you have to make it make features very newcomer friendly let's say by not uh bombing them with a lot of fields because as a newcomer like when i was a newcomer i did not have much to show much to talk about so you know try to make it a bit more beginner friendly like you know beginner like like questions you don't ask them about the projects they do but ask them but like what they want to do and you know other like how they how do they want to go follow types because they will be their main audience sure we will uh try to implement those terms and uh we have uh done this kind of like implemented many chat section also to talk so so that people can actually have conversation about their projects or something like that so we will surely uh keep in that mind that it should be like for the beginner friendly also so yeah anymore and uh you look forward for the your three communities because they haven't come to happen right so this your product could be actually like you know using me in a mainstream if you cater to the particular audience and the web3 audience is a very easy is an easy catch right now beginners decentralized petitioner by team binary brains and i use shaw will present i wish you may stay on your screen so can we leave yeah yeah we can okay thank you oh hello just a minute i'm sharing the screen sure would give me some one minute i'm okay with the next project next we have next we have nirvaya by push push i think someone else is okay hi so we are crazy days and we are here with the solution which is a mainly solution remove to avail health care items and remove the social coordinates next the main problem that we are solving here is that the unavailability of the contraceptions and you know emergencies and different places and remote places where we can where converted pregnancies and different health issues can be avoided and the second problem that we are solving is the unhealthy unavailability of sanitary pads in the emergency sequence for women and also promoting their use and this um and destroying the cabin now if we come to the main motivation behind our project okay you can say the things which have inspired us to make this project are is firstly the feeling of shame that people cannot ask for contraception and such kind of things in tier 2 tier 3 but even in some tier 1 cities and sometimes this also happens that people get these items through delivery either when you need those in emergency you cannot order it two to three days earlier and when you take taken with 10 minutes or 20 minutes or something like that firstly it takes a large cost and second the products may not keep the privacy that you know thirdly is that availability in many village or tier 2 tier 3 tier 3 cities especially we see the medical stores are not very close by to our house or not in a walking distance so in the case of an emergency or in the case of any kind of needs we have to go either we have to walk or we have to rush from medical school hospital or some places like that buy such things and last thing is the table that i am working that we are we have noticed is that especially the women in the world of the backward areas are still using primitive methods like etc for menstruations and for all others kind of things like that okay so through this we can promote a very very healthy healthy uh birth control as well as some problems related to shelter disease and all sorts of things now we have to design a iot based system and hardware along with an app where the user can see the availability of the products and the machines around the location and we can just and he or she can select the nearest machines or the machine that he likes you just play with the basis of pay from the app with the basis of upi and collect the required items and this is as it is a web app so we can say that the machines are having a entire database of things which machine has how many number of things present with it and some things like that and we can also manage this centrally and the last thing is that to prevent illegal business or buying through machines we have used a red limiter so that a person can buy a certain amount of things within a certain period of time now for how the how the app works and a live demo of the hardware that we are making i am [Music] foreign [Music] so of the features we have mentioned how many features are have been implemented during the hackathon we have actually in the harder part like a sort of vending machine is some harder to uh dispense the item that is not in that but the communication between our system and the harder part has been done [Music] dispensing system hardware is not available to us yeah that makes sense i don't expect you guys to have a better machine today so so what you're doing is you're allowing people to buy okay you're you're allowing villagers yeah just promoting safe sex and means you know protection for women during the mess duration okay so like going forward one problem would which i say in your business case would be installing the vending machines and places because i don't i don't think that we have such a culture in india right in india we see that people are looking at games so vending machines are very small sort of things so we plan to if we are going to implement this we plan to implement it with the government through the government okay or some sort of angels that is what it was give a little kind of subsidy and also regulate the quality of the product and also you know take the data afterwards how many what is being used who are you trying to use the kids like who are you targeting uh frankly we are target firstly for the sanitary napkins and they are targeting the women who are means you know they came on road and they just forgot and they are just having their things something like that and and the people who you know are dire need of a contraception in some kind of situations and also in case of if we consider the case of rural cases often safe sex is not promoted in such places so we go through our machines and through government schemes or something like that we can just promote it so what i see look first of all good job connecting the hardware to the software i think the lawyer the only team who has had a hardware component in their project which yesterday and secondly the main problem which i see going forward with the project is you know it's convincing the public otherwise and marketing and throwing the christmas basically right yes our main target is to remove the awkwardness of or you know the taboo of uh talking about sex promoting safe sex something like that but like i mean so so if it's a vending machine it won't be any different from a medical store right like an equal to medical store but uh you don't have to walk to hitting a you don't have to walk some distance for a medical school and you know uh mainly if we if we just consider the tier two and the tier three cities which mainly are in india and the villages yeah the medical software at a very very large distances i am from a th i am from a tier 2 city you can call it so oh it was a it happened that i won't share my stories like that there because [Music] okay and like how you are communicating like how the hardware is communicating to the payment like which particular protocol you are using uh no for all the the server and the communication between the hardware is how many order is received and how an order is dispensing the now the payment services are fully implemented in server side now when someone in the app place the order uh we created an order in our payment gateway site and redirect the user to the actual link when the user payment we have a way for calls and uh they are it says database that yeah yeah actually the all machines are the specific missing code so uh when which machine it needs to the uh as a distance how many items it actually stores in database and the next time the machine actually holds the server then receives the data and displays the items actually will uh maintain a cue like structure in database okay okay i got it i got it nice like one more thing like let's say uh one particular uses uh one particular user let's say uh places an order through your vending machine so is there a unique code id generated for each and every like purchase so that whenever the person goes to the vending machine uh so they can like identify the particular proper person is only getting the product they want uh actually actually we have not implemented that we are actually restricting user to i mean 100 meter 100 or 150 meters of range and so the uh so the actually vending machine available uh it up to two kilometer for the demonstration purpose we have keep the actual distance either we have calculated the euclid uh distance and between the coordinates and filter on basis of that actually so it don't allow to order under 100km or 150k great idea by the way any more questions i don't questions thank you thank you team next we have humanity decentralized petitioner by ayush i wish you can present hello yeah hello you are not audible you can unmute yourself if you can unmute yourself [Music] well okay uh let's move you don't know that when you can fix your resources okay okay it's fine it's fine you can present okay well this is womanity web 3 project we are more of a petitioner and an online boarding website concerning women empowerment our mission is to create a platform where women can demand the laws and regulations they want in a democratic the project is made using travel framework [Music] [Music] so yeah yeah you are please uh be fast it's an online boarding website and more of a petition where we can create petitions submitted and the and the other people can upvote or vote against a petition well this is a petition website concerning women's security our track is web3 and women's security so if i can if i want to upload a web proposal i can write anything like ban marriages it's a simple random proposal which i want to say six yeah so backward is very fast yeah so you can submit here then metamask transaction will come we need to confirm it and then we are seeing there is a new proposal coming banned child marriages we can go into it vote for it for and then again confirm the transaction using meta mask [Music] the whole thing is hosted in robson test network of ethereum after the transaction gets confirmed we can see the boards changing here like we can see votes for one we can come back and even vote on other petitions [Music] we can see that here there are other petitions i have already voted in this part and there is a feature in it that we cannot vote in a partition more than once so if i can if i vote here then the transaction will most probably fail like we can see a warning so then a transaction will fail here because we have already voted and we will see a warning like we can you have seen the transaction has failed the main reason being they have already voted in this so let's come back in this way we can create proposal and show the government that so many people are supporting a proposal so the government can take actions again for concerning the matter and in this way we can bring change for women's security okay thank you very much okay you have anything nice yeah we have implemented the project using truffle framework web3.js library we have used wifi yeah we have used react for the content and that's all i can find so sebastian's do we have time for peer uh q a uh yeah you may give one feedback okay i'm gonna be very brief i'm gonna be very honest right now take care yeah good job with the ui and the whole playing touch and buffer touch to it nice round boxes right but honestly when i come to the smart contract part what i felt was you just rebranded the basic voting app into something for human and that's how into something for women's security and this spacing this basic boating app is a youtube tutorial so i do not see much of an innovation over here and uh i would want to know about the issues about the original i feel like this is more of a backup idea right now and i would want to know the emotional idea which you were working on and the troubles you faced while working on it because what i see right now is just a rebranded version of the plain portfolio yeah well what issues do you face when we wanted to we first of all thought about making up more of a petitioning app then there was a track woman concerning women safety so we thought can make it as a theme concerning women's safety of this main app so your original idea was a simple partitioning app yeah a backboarding system just normal partition right yeah good job okay thank you next we have project travel by aditya krishnan you may present your screen yeah you can leave please stop presenting am i audible yeah okay cool yeah you must share the screen yeah so is my screen visible now it is it is yeah okay cool uh so hi guys uh it's been late night so i've been introducing my project it's called travel uh travel is a your smart travel buddy it's powered by uh artificial intelligence in some part of the functionalities so it basically said travel again uh a backup for your travel plans like we all travel a lot throughout our job so throughout our everyday needs so this application will come handy in certain places list the following and the problems uh we felt uh on a daily travel journey uh or a tourist journey the first thing will be obviously the language barrier so a language becomes the most major barrier in terms of a foreign tourist or a person who is going to visit a native place so he may not be able to um recognize or speak in that particular native language so that is the first problem and the second is uh the market so local artisans and craftsmen uh may not have access uh to the global audience so that is a critical part and next is the availability of a trustworthy gate maybe so a lot of places where you go you may not get a trust for the gate and sometimes these uh gates may offer you cheap products in terms of in return of a large amount of money so these are the major problems and the solution we have is a language translator which is backed by google translate api and the next one is a native market framework where these local artisans and craftsmen can list their products in our framework and like they get to the people get to see like a global audience get to see uh the whole thing uh and the next is an ai powered travel gate which works on the basis of a machine learning model where you can detect a monument or similar places uh in terms of a pre-trained model which is fed into the mlt flight model so initially we have an app and we also have a website running parallel uh serving similar purposes and so talking about the product demo the first thing i would like to share is a scan monument feature though so a scan monument feature uh in multiple times we go to a certain place and we wonder what uh what function or what significance does this thing have um with respect to the particular historical or hierarchical importance of that particular place so uh scan monument feature basically gives you an option to click an image of the peak point of that monument and the ml model automatically recognizes which monument is that and it also gives you the historic behind hierarchical details uh in the following and uh sometimes you might not be in a good mood or you might not want to read so it also has a text to speech engine built into the system and it can also uh like read out the whole thing for you so this is the first scan monument feature and the second is speak native speak native basically is a voice to voice translator which works on the basis of a google translate api and this might have come handy when you go to a native place and you are not able to speak that particular language so you can use the help of this language translator uh in terms of your speaking needs so this is the second and the next is a travel card where you can leverage your uh daily outgoing trips uh in terms of vip uh purchases of vip stays when you hit a particular milestone in the travel car milestone so when you travel particular points or particular locations uh you are you like you get or you accumulate points and these points you can redeem uh when you hit the milestone so this is basically travel current the next is a native market as i said were these local artists that craftsmen can enlist their product into the global audience so the next time a tourist comes to your native place uh he can like recognize unique products so native products these people make and they can buy them from those particular places with trusted authenticity tested by our team so uh the next the final two features will be a suggest a plan method a suggested plan basically what it means is you go to a place and the only thing you know might like uh i personally uh go to a location and uh just for seeing visiting uh just uh just one place oh i'm only aware of that place in that locality and i go there so suggest a plan what it does it gives you a path through which you can go and the path will cover like uh the less significant or uh tourist places which you may not know through the path and this will also suggest how much distance you will cover and how much time this the whole adventure will take through this path and the suggested plan method is also backed up by a feedback system so if the previous uh tourists have gone through this path he will give a rating system and uh the plan a plan we plan to and these plans will be sorted according to the ratings so if the previous visitor have liked it he will also give a good rating and the next tourist will also follow the same path and the next is a locate native feature where the gps takes in your current location and it feeds you with the nearby restaurants nearby auto taxi stands so when you are in a remote location you don't get lost in hey are they just sorry to interrupt you yeah i'll be winding up like two uh two three minutes so yeah uh this is the market size when we analyze like uh 51.9 percent uh it goes to the kerala tourism like we have taken a case study in respect to kerala where i am and 51.79 percent goes into the kerala tourism in increase in growth and three percent grow to the information sector we focus on or we've been point on and the 1.5 percent revenue for private organization like uh what we are planning to build writer so how we are planning to make this thing to the next level make this thing into a business model is like we will be initially having a freemium model where people will uh go on test the app like they'll be given free tries of scan monument feature or any other feature and the next step is like if he or she likes the feature he will be like we will ask for a subscription plan uh in terms of a freemium model so there will also be a small amount of ad revenue from uh this local companies with respect to the specific locality so they also get a kind of promotion and we will be also we partner partnering with tourist entities in that sector itself and also we will be taking commission from the local vendors so under the hood uh things were inside of flutter app uh basically used for mobile application development and we have a tf lite model tensorflow model for image classification machine learning for the web we have a react js and for backend and host servers we have node.js uh so the traction of milestones uh we are uh we have a minimum mvp right now and we are planning uh to deploy the thing and market a bit and we'll be attaining 25 000 user base and the next phase will be a target revenue of 2 lakhs so yeah guys uh that's a wrap and i'll be quickly showing you a demo it's visible right my screen yes it is yeah okay yeah so here it is so this is the travel application uh it will fetch the current location and you have uh like uh options like speak native where you get uh uh voice to voice translation and you can also get a text to uh voice translation of course this is a speak native feature and the next one as i said is a scan monument feature where you can take your image from a camera or gallery and like if you give this i like i will take a pic from uh my laptop and it automatically recognizes that this thing is a also gives the historical hierarchy button and can also read the thing out with the help of a tts engine so the next thing is locate native where the current location is stretched and the nearest restaurants and auto taxi stands are all so fetched into the software and the suggest plan method as i said recommends a path through which you can travel where you cover more places in terms of less time on less distance so you have multiple plants out there and you'll also have a travel car which is also a unique selling point where um in terms of how much you travel points are accumulated and at the end of the day you can leverage this thing uh for vip purchases and vip goods uh so yeah and we also have an 80 market uh currently we don't have items in there and this is the whole thing so i'll be suddenly winding up yeah okay thanks thank you aditya questions i mean there's one question from my side like uh the kind of like like you're kind of creating a suggested path for the user through which they can move so is it just like uh like um ai generated one or it's like uh you know not like what i'm trying to say yeah okay yeah i'll give a brief uh details of what the question like i look like uh so uh we basically have uh some common parts uh in particular localities through which we can go and we cover a lot of uh unique places which we might not be knowing because we are from a different location so we will go through those path and visit these places which aren't in our consideration so we yeah so we have this parts in this and the previous user will obviously give a feedback according to how he felt the place was and current user will also get to know that this place is good and the model will also suggest a path according to how the feedback is so then the next user get to choose what he needs according to the feedback offer i would suggest like you can you could have added some of the like uh ups of various you can say traveling agencies like ola over where they like show up the cost of like if they travel to here to here what's the estimated cost it might take them so you you could have highlighted that as well yeah the suggested plan features how much distance you cut through the uh through the distance or the through those uh monuments it obviously suggests the distance and it also suggests that time how much time an average person will use through those places three or four places through this distance uh so yeah we will it will also be nice if uh how much price we expect to be used in those places it will also be a good feature to add it any more feedback okay thank you yeah thank you guys next we move to the last project of this session with smart passion monitoring system by a team i see s warriors and it will be presented by jagan jagan you can share this it thank you this is smart patient monitoring system our team it's our uh my name is mark our project is to describe in the modern world how the patients are suffering from to visit doctor and doctors have been able to see the past history of the patients we have uh introduced a new idea method with their android app development uh our idea is to make the patients about their history with the past and present in technology we have used modemcu and gy9r6 contactless temperature sensor and max3010 sensor this is our pcb design for uh contactless temperature and uh max three zero one zero three one double zero sensor on the transmitter our main motor is to uh don't affect the patients with the direct contact with the doctor if it is the device which is used to help the patient and doctors without contact they can able to send the data via the node mcu if we kept the device near the patient it can collect the information and deliver to the doctor without seeing the patient the patient the doctor could disable their data anywhere anytime in any places user case electronic health records patient portals we can made it to app the use of technology for embracing uh this is our first project sir uh we couldn't develop us so much learn more thing for that to the projection so thank you thank you jargon the team name is ics warrior amman you are asking yes yeah so any more any questions so like the pcb is it it's a in itself the microcontroller or like it is just the sensor itself only sir uh so for what what uh both of us are it's for a micro nodemcu three to eight three nodemcu and the uh for three sensor set with oil we can able to see the output in violence so like which you are using node mcu 32 yes and not mc32 it is a dual core we can use it uh bluetooth or wi-fi okay so i think so like uh with bluetooth using the normal ble or the low uh low energy ble so now serving i used only wi-fi for this sir we are not uploading the bluetooth okay okay we also for this we created a local ip address set okay after that we created html page and we can able to see in a html placeholder for output of the senses so like are you aware of the fact that you can also embed a website on your nodemcu sir uh for that one local ap addresses okay for uh i have to interface the two codes here i have little experience no no no i i mean i don't i was not talking about the local list i was actually talking about the fact that if you are aware that a nodemcu has a dual core processor and like we have some like 4mb amount of space within it where you can like easily host like website in the json format itself so like why didn't you use that for the website and you like went for a separate website so uh so the question is why we use a normal nodemcu what so can you repeat the question sir i can't understand the questions uh i was asking like uh in nodemcu actually you can embed a website in the json format i mean you have to write the code for the whole website in the json format and in that way you didn't need to like make a separate localhost kind of website for that purpose so why didn't you go for that like it could have been a bit more optimized i mean the idea is good everything is fine but yeah i think so that was kind of an improvement that you could have brought so yes sir sir we have uh so first project sir it's interesting okay we have first step into this uh further we have to learn many things about this project sir okay so do check out this library called arduino json i mean if you might have heard about it before i don't know if you might have heard about it or not yes but i did it i didn't guess that but i i will do so thank you for information yeah thank you again yes thank you any more questions okay okay so thank you everyone thank you judges for staying up so late we know it was there was a delay from our side and we're really sorry for that and also we're thankful to all of the judges and parties friends for the time they gave us for the support they gave us to webv's nitr for the support they gave to is a 2.0 so so thank you with this we are officially towards the end of acquisition 2.2 it was a really wonderful event and we hope everyone gained a lot from it and found it really great thank you and any doubts anything you can just dm any of the organizers on discord will be actively looking forward to your dms thank you very much thank you thank you so much yeah okay bye guys you