Deploy and Host Workmates on Railway

Workmates gives you a persistent workspace where multiple AI agents can work independently or collaborate as a team. Agents can have their own roles, skills, tools, projects, and conversations while sharing rooms for coordinated work.

This Railway template runs the Workmates server and web application as a single cloud-hosted workspace that you can access from a browser, mobile device, or the Workmates desktop app.

About Hosting Workmates

Running Workmates on Railway turns the normally local workspace into an always-available cloud environment.

The deployment includes the Workmates API server and web/PWA interface, with persistent storage mounted at /data. Agents, conversations, installed skills, workflows, configuration, uploaded files, and activity records remain available across restarts and redeployments.

Railway also provides the public HTTPS endpoint used by Workmates clients. Access to the workspace is protected by WORKMATES_SERVER_TOKEN.

After deployment, open the generated Railway domain, authenticate with the server token, and connect an AI provider such as OpenRouter, Centra, or another supported Workmates engine.

Common Use Cases

Run a persistent team of specialized AI agents from one workspace.
Coordinate multiple agents in shared rooms with structured turns and handoffs.
Access the same agents, conversations, files, and workflows from desktop and mobile.
Build agent workflows using skills, schedules, triggers, tools, and approval steps.
Connect external services, MCP servers, browsers, and compatible AI providers.
Keep agent activity and permission requests visible through Workmates' action and approval interfaces.

Dependencies for Workmates Hosting

Workmates requires an AI inference provider or compatible agent engine before agents can generate responses. Supported options include OpenRouter, Centra, custom OpenAI-compatible endpoints, and other engines available through Workmates.

The hosted workspace also requires persistent storage. The Railway template mounts a volume at /data so workspace state survives application restarts and redeployments.

The cloud image contains both the Workmates server and web client, so no separate frontend deployment is required.

Variable	Description
PORT	Port used by the Workmates HTTP server. Railway normally provides this value automatically.
WORKMATES_HOSTED	Enables Workmates cloud-hosted mode for deployments running on Railway.
WORKMATES_DATA_DIR	Directory used for persistent Workmates workspace data. Set to `/data` when using the Railway volume.
WORKMATES_STATIC_DIR	Directory containing the built Workmates web application served by the hosted server.
WORKMATES_SERVER_TOKEN	Secret access token used to authenticate users and Workmates clients connecting to this workspace. Keep this value private.

Implementation Details

The template runs the published Workmates cloud container and exposes the application through Railway's generated HTTPS domain.

Persistent workspace data is stored at:

/data

The deployment should use a single Workmates server replica because the workspace currently uses file-backed persistent state.

Once deployed:

Open the Workmates service in Railway.
Copy the generated WORKMATES_SERVER_TOKEN.
Open the Railway-provided public URL.
Sign in using the server token.
Connect an AI provider from Workmates Settings.
Create agents and begin using the workspace.

The same Railway-hosted workspace can also be opened from the Workmates desktop application or installed as a PWA on supported mobile devices.

Why Deploy Workmates 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 Workmates 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.