Insightly

AI meeting transcription and analysis: live recording, file upload, transcription (Deepgram), AI summaries (You.com), and PDF reports (Foxit). Built with React, Flask, MongoDB, and Docker.

---

Quick Start (Docker)

Prerequisites: Docker and Docker Compose

# 1. Clone and enter project
git clone <repo-url>
cd devweekhackathonFeb2026

# 2. Configure (optional — works with defaults)
cp .env.example .env
# Edit .env: set DEEPGRAM_API_KEY for live transcription

# 3. Run
docker compose up --build

Open http://localhost:5173 — sign up, then record or upload meetings.

---

Configuration

Copy .env.example to .env at the project root and edit as needed.

Variable	Required	Purpose
MONGODB_URI	Yes*	MongoDB connection. Default mongodb://mongo:27017 works with Docker. Use mongodb+srv://... for Atlas.
DEEPGRAM_API_KEY	Yes**	Deepgram — live mic + file transcription
YOUCOM_API_KEY	No	You.com — AI summaries (insights, decisions, action items)
FOXIT_CLIENT_ID / FOXIT_CLIENT_SECRET	No	Foxit — PDF meeting report generation
VITE_API_URL / VITE_SOCKET_URL	Production	Backend URL reachable from browser (default: http://localhost:5000)

* Docker default works out of the box. For Atlas: set MONGODB_URI and run docker compose up --build backend frontend (omit mongo).
** App runs without it, but live mic and file transcription won’t work.

---

Deployment Options

Option A: Docker (recommended)

docker compose up --build

Runs MongoDB, backend (port 5000), and frontend (port 5173). All config via root .env.

MongoDB Atlas: Set MONGODB_URI in .env, then:

docker compose up --build --scale mongo=0

Option B: Local (no Docker)

Prerequisites: Python 3.10+, Node.js 18+, MongoDB, ffmpeg

1. MongoDB — Local or Atlas
2. Backend:

   cd backend
   cp .env.example .env   # Set MONGODB_URI, DEEPGRAM_API_KEY
   python -m venv .venv
   .venv\Scripts\activate   # Windows
   pip install -r requirements.txt
   python app.py            # → http://localhost:5000

3. Frontend (new terminal):

   cd frontend
   cp .env.example .env     # Usually empty for local proxy
   npm install
   npm run dev              # → http://localhost:5173

---

Troubleshooting

Issue	Fix
"DEEPGRAM_API_KEY must be set"	Add key to root .env (Docker) or backend/.env (local). Restart.
MongoDB connection errors	Check MONGODB_URI. Ensure MongoDB is running (local or Atlas reachable).
Socket/API connection failed	Backend must be on port 5000. For remote hosts, set VITE_API_URL and VITE_SOCKET_URL in root .env.
Microphone permission denied	Use HTTPS or localhost — browsers block getUserMedia on plain HTTP.
MP3 download fails	ffmpeg required. Docker image includes it.
PDF report 503	Set FOXIT_CLIENT_ID and FOXIT_CLIENT_SECRET. Check /health for foxit_configured: true.

---

Project Structure

├── backend/          # Flask, Socket.IO, REST API, Deepgram, Foxit, You.com
├── frontend/         # React (Vite), Socket.IO client, Web Audio API
├── docker-compose.yml
├── .env.example      # Copy to .env for Docker config
└── README.md

---

Tech Stack

• Frontend: React (Vite), Tailwind, Radix UI, Socket.IO, Web Audio API
• Backend: Flask, Flask-SocketIO, MongoDB, GridFS, JWT auth
• APIs: Deepgram (transcription), You.com (summaries), Foxit (PDF reports)
• Audio: WebM storage; ffmpeg for MP3 export