Back to projects
Multi-Agent Transcript System
Transcript processing system built on multi-agent workflows in LangChain and n8n automation, with a FastAPI backend, MongoDB storage and a React interface.
- Project
- Applied AI project
- Role
- Agents, backend & UI
- Period
- 2024
[ Overview ]
The problem
Reviewing long meeting transcripts by hand to pull out summaries and follow-ups is slow and inconsistent.
What I built
Split transcript processing across cooperating LangChain agents, automated the surrounding workflow with n8n, and exposed results through a FastAPI backend and React interface.
[ Highlights ]
- Multi-agent workflows in LangChain, each agent handling one part of transcript processing
- Workflow automation with n8n
- FastAPI backend with MongoDB for storage
- React interface for reviewing processed transcripts
[ Stack ]
- LangChain
- n8n
- Python
- FastAPI
- MongoDB
- React