Deploy, scale, and manage AI agents with sleep/wake architecture.
No infrastructure. No surprise bills. Just ship.
Website · Docs · API Reference · Blog · Discord
Traditional cloud wasn't built for AI agents. Serverless functions time out. Containers waste money sitting idle. VMs require DevOps expertise. Maritime changes this with a purpose-built runtime that understands how agents actually work.
Sleep/Wake Architecture — Agents automatically sleep when idle and wake on-demand in under a second. You only pay for actual compute, not idle time.
One-Click Deploy — Connect your GitHub repo. We handle the build, deploy, SSL, and routing. No Dockerfiles, no YAML, no CI/CD pipelines to maintain.
Built-in Triggers — Invoke agents via cron schedules, webhooks, REST API, or messaging integrations. Everything you need to orchestrate agent workflows, out of the box.
Connect repo → One-click deploy → Agent goes live → Sleeps when idle → Wakes on trigger
Every agent gets a stable HTTPS endpoint, real-time log streaming, environment secret management, and a full lifecycle API. Deploy agents built with CrewAI, LangGraph, OpenClaw, or any custom framework.
| maritime-sh/cli | The Maritime CLI — deploy and manage agents from your terminal. |
