Epstein Transcript Explorer

Transcription, search, and Q&A over the Epstein-Ehud recording. Uses Deepgram for transcription with speaker diarization and entity detection, a Next.js frontend for exploration, and Modal for deployment.

Live demo: https://nazihkalo--epstein-transcript-explorer-serve.modal.run/

Demo

epstein_dmeo.mov

Setup

1. Install Python deps

uv sync

2. Configure API keys

cp .env.example .env
# Edit .env with your DEEPGRAM_API_KEY and OPENAI_API_KEY

3. Run transcription

uv run python main.py

This reads the MP4, sends it to Deepgram, and writes:

• data/transcription.json — raw Deepgram response
• data/transcript_structured.json — cleaned structured data

4. Local development

Run the backend and frontend in two terminals:

# Terminal 1: FastAPI backend on :8000
uv run uvicorn server:app --reload --port 8000

# Terminal 2: Next.js dev server on :3000
cd frontend
npm install
npm run dev

Open http://localhost:3000 — API calls are proxied to the backend automatically.

5. Build frontend (for deployment)

cd frontend
npm run build
cd ..

6. Set Modal secrets

modal secret create openai-secret OPENAI_API_KEY=your_key_here

7. Deploy to Modal

# Development (hot reload)
modal serve modal_app.py

# Production
modal deploy modal_app.py

Architecture

• main.py — Deepgram transcription pipeline (run once)
• server.py — Local FastAPI dev server (same endpoints as Modal)
• modal_app.py — FastAPI backend + static file serving on Modal
• frontend/ — Next.js app (static export)
• data/ — Source MP4 + generated transcript JSON

API Endpoints

• GET /api/transcript — Full structured transcript
• GET /api/search?q=...&speaker=... — Search with optional speaker filter
• POST /api/ask — Q&A with RAG over transcript (requires OpenAI key)