Technical
Runbook — local dev
Start everything, reseed, and run the AI host.
Start the stack
# from repo root
docker compose up -d # postgres + minio
cd api && source .venv/bin/activate
alembic upgrade head
uvicorn app.main:app --reload --port 8000
# in a second terminal
cd web && npm run dev # binds to :3010
Reseed the demo
# DROP and rebuild the dev DB, then reseed
psql -h localhost -U postgres -d teloma -c "DROP SCHEMA public CASCADE; CREATE SCHEMA public;"
cd api && alembic upgrade head && python -m app.seed
Research AI
The Research Assistant is powered by Calanthe (NxGen AI Labs' flagship AI), running on Azure OpenAI as the primary path. An optional local Ollama model is supported as a secondary fallback, and a deterministic template restatement is the final fallback — so the assistant always returns something safe even if the AI hosts are unreachable.
Note
In production, Teloma runs on Azure: the web app on Azure Static Web Apps, the API on an Azure Container App, Postgres on Neon, and documents/images in Azure Blob storage (Microsoft BAA). Local dev uses Dockerized PostgreSQL 16 + MinIO.