Meeting Insights AI
An AI meeting-intelligence system that turns transcripts into searchable team insights.
The challenge
Teams often lose valuable information inside long meeting transcripts, scattered notes, and shared files. Finding decisions, action items, blockers, and previous discussion context can take too much time. The system needed to securely store meeting knowledge, make it searchable, and generate useful summaries and insights automatically.
What 13Floors built
13Floors built an n8n automation workflow that captures meeting transcripts, stores them in Google Drive, extracts and cleans transcript content, creates embeddings, and indexes the data inside a Qdrant vector database. I connected OpenAI and Gemini models through a RAG pipeline so users could retrieve relevant transcript context and generate summaries, decisions, action items, risks, and follow-up insights.
The value created
The system helped teams turn meeting transcripts into a searchable knowledge base. Users could quickly find past decisions, extract action items, understand project context, and make better data-driven decisions without manually reviewing long meeting records.
Technology
- n8n
- OpenAI API
- Gemini API
- Qdrant
- Google Drive API
- RAG Pipeline
- Vector Database
- AI Agent
- Webhook Automation
- JavaScript