Google has been on fire with the release of Gemini 2.0 model series these models are built specifically for AI gentic use case Google even markets them as tools for creating AI agents and automating things with it within the Gemini 2.0 model family you have four models all designed for gentic tasks the 2.0 pro which is an experimental is best for coding and handling complex prompts the 2.0 flash is fast powerful and optim for gentic AI the 2.0 flash thinking experimental model is used for enhanced reasoning and something with explainability the 2.0 flashlight is the most cost-efficient model beyond that these models offer multimodal support which includes image generation text to speech as well as tool use which is going to allow agents to utilize different sorts of things like Google search executing code and so much more the 2.0 Flash model in fact even supports a 1 million token context window which is making it practical for everyday use by referencing lengthy documents and processing large data sets as for pricing these models are free to use with the free tier which is insane with the Gemini models you can build AI agents with built-in agenda capabilities for free faster than any model and with ease and that's exactly what I'll be showcasing throughout today's video as I'll be using Vector shift to help you create AI agents with powerful automation capabilities powered by these Gemini models for people who are unfamiliar Vector shift is an all-in-one AI platform that allows you to build AI apps automations through a simple drag and drop interface it's incredibly easy to get started and best of all it's free so let's get straight to it go over to Vector shift's site and click on get started you can either log in if you have an account or you can create one with an email address or with Google or with GitHub once you sign in you're going to be then sent over to the main dashboard of vector shift this is where you can manage all your automations chat Bots and so much more on the left hand panel you can manage everything from knowledge your files the interface in which you can manage your Creations from chat Bots to voice Bots you have the ability to analyze the performance of your agents and you can even do that through the evaluation but what we're going to do is create a pipeline from scratch so that I can showcase how you can create AI agents within this platform we're going to go ahead and create a new Pipeline and what you'll notice is that there's going to be readymade templates that you can already access you create a chatot template or something like a Blog article generator you can essentially practically automate things through different Integrations like Gmail like where you can automate your emails as well as slack messages and so much more but in our case let's go ahead and create a pipeline from from scratch this is the main dashboard of vector shift this is the drag and drop Builder and essentially every workflow needs an input node and an output node to be quite functional now with today's Vector shift example of creating an AI agent I'll show how you can use the Gemini 2.0 model that will help automate repetitive tasks using the list mode that Vector shift has this is where you can run procedures multiple times to handle mundane tasks and when you have it powered by Gemini's multimodal support speed and large context window the AI agents can process more data more informed decisions can be made and manage complex workflows quite efficiently now to demonstrate this we'll build a lead classification workflow within Vector shift imagine you're an Outreach rep researching leads manually Gathering contact information about different types of law firms websites data from multiple cities that would take hours instead I'll show you how you can automate this entire process with an AI agent that's autonomously going to find and extract your information like the type of law firm the city it's in the information about it which will save you so much more time and this will highlight how Vector shift plus Gemini 2.0 can EMP power AI agents to handle complex tasks effortlessly so now head back to your pipeline dashboard and head over to Integrations and place down a Google sheet integration node the reason why is because we're going to be outputting all the data to this but say if you're trying to create an AI agent you can easily work with so many different types of thirdparty applications that can help you automate things on different platforms like sending out emails scheduling appointments working with your docs and so much more and in our case we're going to be working with Google Sheets because we're going to have the leads the data being processed from it to help us automate this whole process so what we're going to do is have it so that it reads from our Google sheet that we have created we're going to then have it so that our accounts linked up so make sure you have an account fully linked up to this once linked you can then go ahead and pick a file so we're going to go and pick one of the files that we just created for this particular example and we're going to then have it so that we can select the current sheet that we have which is searches once that is done we can then go over and click on confirm now once you have connected your sheet you're going to then see that the output field is operational and it is now populated now to have this fully functional we're going to need to actually create another pipeline it's essentially a sub pipeline that will make the main pipeline operational to collect the necessary Fields so go ahead and create a new pipeline from scratch now this sub pipeline is going to be helping us analyze the URLs to fill out the certain fields that we have within the Google sheet so go ahead and place down two nodes one is going to be to help us process the URL itself so go ahead and rename this to URL now the second node is going to be for the Google searches to fill out the search tab that we have over here next you want to head over to data loaders and you want to place down a URL web scraping node this will scrape all the URLs that we have so you can go ahead and connect the first URL node and essentially this is where every time we go over to find leads on Google this will take the process of going through multiple pages with the recursive feature and it will automate finding different leads for you on the multiple Pages next to power all this we're going to go over to the large Lish model node and place down two Google noes and this is where Gemini tono comes into play we're going to then provide a system prompt for one of the nodes I told it to extract the company about us and then the other one I told it collect the law firm name and now you can power it with the Gemini 2.0 Flash model which is something that I definitely recommend so we have it so that it will extract all the necessary things that we have over here but now we're going to need to place down another Google sheet node so that it outputs it to this table over here that is named data so go over and click on add new row we're going to then actually zoom in so you can see a little bit better and then click on next then pick the file that we have that we've just created and make sure it's on the right tab so in this case it's on data so make sure you select data then confirm the selection and now for the fields that we have or wanted to Output so for the law name that we're trying to extract make sure you select the correct one click on this plus sign and select the correct node and then you wanted to also select the output as responses now do the exact same thing for the aboutus which is the Google Zero node so click on the plus sign click on Google zero and then responses and then lastly this search node the input that we have make sure you click on the plus sign and then select the search and then have it so that it is text now lastly you just need the web scraper connected to the large language model so click on this plus button and then you want to then select the URL loader and then you want to have it so that it is going to select the content output and do the exact same thing for the other large sish model now go ahead and click on deploy changes and you can go ahead and click deploy now what we're going to do is focus on the next pipeline so you want to click on create a new Pipeline and the reason why we're doing this is because we're focusing on the layer above this this pipeline will just focus on the Google search and then call the other pipeline that we just created to do the other things like extracting and do the recursive searching place down an input node and then renaming this to Google search and then we just need to place down a web search node so head over to data loaders and place down this web search node and then select Google search and then we can connect it to this node and then say for the results we wanted to find the top five results lastly we just needed to call the last pipeline we created so head over to General place down this pipeline node and then go over and select the pipeline that we just created then what you want to do is turn on the list mode and what you can do now is then input the Google search node that we have and make sure it's URL then head over to data transformation and you want to then place down a list Ops node the reason why is cuz this node takes a list of texts because of the list mode and then this is something that only takes a singular text which is why you will need a list operation node so go and click on duplicate and what you want to do for item for duplication is the initial search so click on Google search which is the first node and then have it so it's outputed as text and then list to match is the Google search which is from the other Pipeline and make sure it's on URL you also need to make sure that you have the output connected to the last pipeline node so make sure you do that by then having it so that you select the duplicate list and then having the output selected as well so now we have all of the sulur connected and fully configured so go and deploy the changes now we need to go back to the master sheet so in the main dashboard make sure you click on the master pipeline now you just need to Simply have it connected to the other pipeline so you don't need an input node or an output node you just need to have the other two pipelines connected to do so go over to General and then place down one Pipeline and make sure you click on enable list mode so add in the Google search for patent lowers and and then for Google search make sure you connect it and make sure you select searches now you can simply just click on deploy changes and what's essentially going to happen is it's going to go search these five contents for each of the three types of patent law firms from New York San Francisco Boston it's going to then process through all of the layers and then output it as data as the law firms the search result as well as the Bas so to test this out what we can do is obviously deploy it and then now you can click on run pipeline just simply click run and then if we go over to this Google Search tab it's going to then output all the law firms it's going to give us a summary and it'll be able to do this every time we click it it'll be able to extract five sources so let's just wait and have it f through all the leads and there we go we have the names of all the law firms as well as the B us page and you can see that gives us a lot of information on this because of what we had requested for in terms of the name as well as what the the actual law fir is about and that's how essentially we were able to automate this whole process we're able to do this with the help of vector shift and then powering it with the Gemini 2.0 models this is just one example of what you can do you can create so many different types of Agents with this so huge huge recommendations to use this I definitely think that this is the best AI automation platform so definitely recommend that you try out Vector shift cuz you can get started completely for free free but with that thought guys thank you guys so much for watching I hope you enjoyed today's video I'll leave all the links in the description below make sure you subscribe to the newsletter make sure you follow me on the patreon make sure you also follow me on Twitter and lastly make sure you guys subscribed turn on notification Bell like this video and please take a look at our previous videos because there's a lot of content that you will truly benefit from but with that thought guys have an amazing day and I'll see you guys early soon peace out f