TITANIC is an AI-native reliability operating system prototype for autonomous incident investigation, predictive reliability, infrastructure intelligence, and recovery workflows.
- Frontend Command Center: https://titanic-sre-engine.vercel.app/
- Backend API: Deployed on Railway (fully configured with dynamic CORS to support the Vercel frontend command center)
apps/web- runnable command center UI with incident RCA, service map, replay, chat, postmortem, and integration settings.apps/api- FastAPI-ready backend scaffold for incidents, topology, AI analysis, chat, recovery, and predictions.docs/ARCHITECTURE.md- scalable product architecture mapped to the blueprint.docs/MANUAL_SETUP.md- required API keys, links, cloud settings, and integration notes.docs/API.md- endpoint contract for incidents, AI, infrastructure, replay, and recovery.infra/docker-compose.yml- local service plan for Postgres, Redis, Neo4j, and Qdrant.apps/api/alembic- SQLAlchemy/Alembic schema migration setup.
From the workspace root:
python -m http.server 4173 -d apps/webOpen:
http://localhost:4173
The current UI is self-contained and uses realistic mock telemetry so the product can be reviewed before real integrations are connected.
The instant demo is no-build for hackathon reliability. The React/Vite/Tailwind migration scaffold is also included in apps/web/package.json, vite.config.ts, and tailwind.config.ts.
After Node dependencies are available:
cd apps/web
npm install
npm run devcd apps/api
.\scripts\setup_backend.ps1
uvicorn app.main:app --reload --port 8000Create apps/api/.env from .env.example before enabling real AI and infrastructure actions.
Run full local infra:
cd infra
docker compose up --build