--- title: "Deploy Open Executive" description: "AI executive team: one persona backed by 8 specialist agents" category: "AI/ML" url: https://railway.com/deploy/open-executive --- # Deploy Open Executive AI executive team: one persona backed by 8 specialist agents **[Deploy Open Executive on Railway](https://railway.com/template/open-executive)** Machine-readable deploy manifest (JSON, validated by TemplateCI): https://railway.com/deploy/open-executive/manifest.json - **Creator:** My Projects - **Category:** AI/ML ## Template content ### openexecutive-api https://raw.githubusercontent.com/ForeactLabs/OpenExecutive/main/packages/ui/src/app/icon.svg - **Image:** ghcr.io/foreactlabs/openexecutive-api:main - **Start command:** `uvicorn openexecutive.api.main:app --host 0.0.0.0 --port 8000` - **Health check:** /health ### openexecutive-ui https://raw.githubusercontent.com/ForeactLabs/OpenExecutive/main/packages/ui/src/app/icon.svg - **Image:** ghcr.io/foreactlabs/openexecutive-ui:main - **Health check:** /signin - **Public domain:** Yes ## Documentation # Deploy and Host Open Executive on Railway [Open Executive](https://github.com/ForeactLabs/OpenExecutive) is an AI-powered virtual executive team: a single coherent executive persona backed by 8 specialist agents, with a FastAPI backend and a Next.js web app. It keeps episodic memory, a people roster, scheduled briefings and a company knowledge base, and can connect to Slack, Telegram, Google Workspace and Microsoft 365. ## About Hosting Open Executive This template runs the two official containers from the project's GitHub Container Registry: - **openexecutive-api**: the FastAPI backend (port 8000), private only, reached by the UI over Railway's private network. - **openexecutive-ui**: the Next.js web app (port 3000) with a public Railway domain. It proxies every API call server side. - **Volume** mounted at `/data` on the API: SQLite episodic memory, ChromaDB vector index, company profile and uploaded documents. - **Generated secrets**: `BACKEND_SHARED_SECRET` (shared by UI and API, gates every API route) and `AUTH_SECRET` (signs the UI session). - **Auth**: the API runs with `OE_PUBLIC_DEPLOYMENT=1`, so it refuses to boot without the shared secret. The UI redirects every page to `/signin`, which accepts only Google or OpenID Connect sign-ins whose verified email is on `ALLOWED_EMAILS` (or the People roster). **After deploy:** 1. In Google Cloud Console create an OAuth client (Web application). Add `https://YOUR-RAILWAY-DOMAIN` as an authorized JavaScript origin and `https://YOUR-RAILWAY-DOMAIN/api/auth/callback/google` as an authorized redirect URI. 2. On the openexecutive-ui service set `AUTH_GOOGLE_ID`, `AUTH_GOOGLE_SECRET` and `ALLOWED_EMAILS` (your email). For SSO instead, set `AUTH_OIDC_ISSUER`, `AUTH_OIDC_ID` and `AUTH_OIDC_SECRET` with the redirect URI `https://YOUR-RAILWAY-DOMAIN/api/auth/callback/oidc`. 3. Make sure `ANTHROPIC_API_KEY` and `EXEC_EMAIL_ADDRESS` are set on openexecutive-api (asked at deploy time). 4. Open `https://YOUR-RAILWAY-DOMAIN`, sign in, and run the onboarding wizard. A "no company profile" message before onboarding is expected. 5. Optional integrations (Slack, Telegram, Google Workspace, Microsoft 365, Honcho) are configured with extra variables on openexecutive-api; see the project's `.env.example`. ## Common Use Cases - A founder or executive delegating briefings, follow-ups and inbox triage to an AI chief of staff - A small company keeping institutional memory: people, decisions, documents and commitments in one place - Scheduled morning and end-of-day briefings and goal reviews - Drafting replies and meeting actions across Gmail, Outlook, Slack and Telegram with human approval ## Dependencies for Open Executive Hosting - Source: [ForeactLabs/OpenExecutive](https://github.com/ForeactLabs/OpenExecutive) (Apache-2.0) - Images: `ghcr.io/foreactlabs/openexecutive-api:main` and `ghcr.io/foreactlabs/openexecutive-ui:main` - Required: an Anthropic API key, the Executive's email address, and a Google OAuth client or an OpenID Connect provider for sign-in ### Deployment Dependencies - [Deployment guide](https://github.com/ForeactLabs/OpenExecutive/blob/main/docs/deployment.md) - [Authentication guide](https://github.com/ForeactLabs/OpenExecutive/blob/main/docs/auth.md) - [Anthropic Console (API keys)](https://console.anthropic.com/) - [Google Cloud Console (OAuth clients)](https://console.cloud.google.com/apis/credentials) ### Implementation Details - The project publishes rolling `main` images (no release tags are on the registry yet), so the template tracks `main`. Pin a `sha-...` tag on both services if you want a frozen version. - The API must run exactly one replica: the scheduler has no leader election. Do not scale it horizontally. - The API has no public domain; the UI reaches it at `http://openexecutive-api.railway.internal:8000` over Railway's private network. - Healthcheck `/health` with a 300 second timeout: a cold start builds the vector index before serving. Give the API at least 2 GB of memory. - `AUTH_URL` is set to the UI's public Railway domain so sign-in redirects return to the right host; update it if you attach a custom domain, together with `BACKEND_ALLOWED_ORIGINS` on the API. - Signed callers (`CALLER_ASSERTION_PRIVATE_KEY` / `CALLER_ASSERTION_PUBLIC_KEYS`) are recommended by the project but need an Ed25519 key pair made with `scripts/make-caller-keys.py`; add them after deploy. - Back up `/data/episodic_memory.db` with `sqlite3 .backup`; it is the part you cannot rebuild. ## Why Deploy Open Executive on Railway? Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it. By deploying Open Executive on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway. ## Similar templates - [Chat Chat](https://railway.com/deploy/-WWW5r) — Chat Chat, your own unified chat and search to AI platform. - [Weaviate](https://railway.com/deploy/weaviate-1) — Vector database with hybrid search and an API key-protected API - [Möbius Personal AI OS](https://railway.com/deploy/mobius-v01-personal-ai-os) — Self-hosted AI agent that builds and runs apps for you, with memory. Open this page in a browser: https://railway.com/deploy/open-executive