How It Works

A standard workflow for deploying an agent to production.

1. Define the Agent

Start by writing a system prompt and defining the tools your agent can use. You can integrate existing REST APIs, Postgres databases, or custom internal services.

2. Deploy the Worker

Push your agent configuration via our CLI or set up a GitHub integration. We automatically provision the necessary edge workers and vector memory stores.

3. Monitor & Scale

Invoke your agent via API and monitor its execution in real-time through our dashboard. Scale up to thousands of concurrent agents without managing a single server.