As junior developers, we often struggled to find answers buried in Confluence pages or scattered documentation. Most of the time, we ended up relying on senior developers for guidance. This was inefficient and slowed us down, inspiring us to create a smarter way to access knowledge.
Our Idea We built a voice-enabled AI assistant that connects directly to Confluence and Context7. Using Retrieval-Augmented Generation (RAG), it delivers instant, conversational answers grounded in real docs. With optional Mintlify integration, it can also auto-generate documentation from user conversations, keeping knowledge bases alive.
Challenges & Learnings We learned how to fetch and parse docs from multiple sources, connect through MCP, and use Pinecone for RAG. Challenges included configuring secure connections, managing latency for voice, and stitching APIs together into one seamless workflow.
Built With
- nextjs
- pincone
- typescript
- vapi
Log in or sign up for Devpost to join the conversation.