Deploy helix-foundry

Deploy and Host Helix Foundry on Railway

Helix Foundry is a self-hosted business analytics workspace for connecting company data, discovering relationships, and exploring metrics through a graph-aware interface. It combines an interactive web UI, a HelixDB graph backend, and a local DuckDB executor so teams can move from raw datasets to an explainable operational workspace.

About Hosting Helix Foundry on Railway

Railway runs the Docker-based application and HelixDB as private services, provides a public HTTPS domain for the app, and supplies managed deployment, networking, and restart behavior.

Tech Stack

• Node.js 24
• HelixDB v0.0.12
• DuckDB executor
• Fastify API server
• Docker images from Docker Hub and GHCR

Why Deploy Helix Foundry on Railway

Railway provides simple Docker image deployment, private service networking, automatic HTTPS, and a browser-accessible URL without requiring source-based builds or local infrastructure.

Common Use Cases

• Explore connected business datasets
• Build graph-aware operational workspaces
• Import sample or customer data
• Query metrics with hosted AI providers
• Review ontology and relationship suggestions

Deployment Notes

The public app listens on container port 3001. HelixDB listens privately on port 8080, while the co-located DuckDB executor uses port 3002. Railway generates encryption and internal authentication secrets automatically. The app service intentionally starts the executor and API together, with the worker enabled inline. This verified deployment does not attach a volume, so workspace data is ephemeral. Local Ollama is not bundled; choose Claude or OpenAI in the browser and enter provider credentials there.

Dependencies for Helix Foundry on Railway

The app depends on HelixDB over Railway's private network. The executor runs inside the app container to keep the deployment compact while preserving the full browser UI and data-processing workflow.

Deployment Dependencies