The meeting provided a detailed, side-by-side comparison of two voice AI platforms: VAP AI and Retail AI.
Key differences were discussed across user interface, cost, voice quality, integrations, analytics, and multilingual support.
Examples were provided for common use cases including agent creation, workflow building, speed, and cost breakdown.
Final recommendations highlighted which platform may suit different business requirements.
Action Items
No specific dated or assigned action items were mentioned in the transcript.
User Interface Comparison
VAP AI: Uses a dashboard with sections like overview, platform, voice library, logs, and agent creation via templates or blank options. Configuration details are divided into separate sections.
Retail AI: Landing page features agents, knowledge base, phone numbers, and more. Agent creation is available via pre-made templates or from scratch. Main functionalities are consolidated into a single page.
Retail AI allows for more customizable welcome messages compared to VAP AI.
Model and Provider Options
Both platforms offer OpenAI and Anthropic models, as well as support for custom LLMs via URL.
Retail AI and VAP AI both allow selection of custom LLMs.
Workflow and Agent Builder Differences
Retail AI: Supports single-prompt and multiprompt agents with prompt trees to manage complex workflows.
VAP AI: Uses single-prompt and block agents, with workflow construction possible by chaining conversation blocks, tool calls, or workflows.
Speed and Latency
Retail AI demonstrated a latency of 1.60 seconds, while VAP AI showed 1.93 seconds in the same scenario.
Voice Quality and Options
Retail AI: Primary voice providers include Play HT, ElevenLabs, and OpenAI. Offers a broad variety of background sound settings and speech customizations.
VAP AI: Supports a wide variety of voice providers and sound customization options (clarity, stability, style, etc.), but with fewer background sound options compared to Retail AI.
Both platforms can use the same providers (e.g., ElevenLabs), but Retail AI was found to have slightly better voice quality from testing.
Cost Analysis
VAP AI: Charges ~$0.11 per minute (using GPT-4o mini model) with a fixed platform fee ($0.05).
Retail AI: Costs ~$0.076 per minute for voice AI (using GPT-4o mini), with a fixed phone number fee of $2/month.
Retail AI is generally more cost-effective for complex agents and higher call volumes.
Integrations and Functions
Retail AI: Has a wider range of built-in functions including call transfer, calendar integration (cal.com), and more.
VAP AI: Similar functionalities but require manual setup via third-party tools; does not have built-in calendar integration.
Testing and Analytics
VAP AI: Offers detailed conversation analytics, success evaluation tools, and structured data extraction with customizable evaluation rubrics.
Retail AI: Simpler analytics with a notable “Test LLM” feature for simulating calls in chatbot mode.
Multilingual Support
Retail AI: Supports 90 languages, demonstrated as the fastest multilingual agent in testing.
VAP AI: Supports 35+ languages, offering broader overall language support.
Final Platform Recommendations
VAP AI is recommended for those needing advanced analytics, more voice options, and customization.
Retail AI is suggested for businesses needing cost efficiency, built-in scheduling, and faster call testing.
Both platforms have unique strengths and the choice should align with specific business needs.
Decisions
No explicit final decisions documented. The meeting concluded with recommendations tailored to different use cases.
Open Questions / Follow-Ups
Participants prompted to comment on their platform preferences or additional questions.
Viewers encouraged to consult additional videos for deeper dives and specific implementation guides.