Turn AI-generated code into shipped software.
Self-serve, or fully managed.
Run the platform yourself, or have us operate it. Either way, every change is verified in a preview before it ships to your cloud.
Self-serve
Free
$0/mo
10 jobs per month
- 1 user
- Role based access control (RBAC)
- Priority email support
- PlanningDescribe an outcome; our planning agent reads your repos and architecture, writes a spec, and decomposes it into Tasks.
- TasksOur agents write code across your repos, open pull requests verified in a preview, and deploy to your cloud.
Self-serve
Team
$79/mo
- Unlimited users
- Role based access control (RBAC)
- Priority email support
- PlanningDescribe an outcome; our planning agent reads your repos and architecture, writes a spec, and decomposes it into Tasks.
- TasksOur agents write code across your repos, open pull requests verified in a preview, and deploy to your cloud.
Managed services
Managed
Done-for-you engineering, operated by our team.
Custom
Annual retainer, scoped to your codebase and Task volume
Includes committed Task throughput + verification
Everything in Team, plus:
- Named engineer who operates the platform for you
- Predictable throughput + verification — we own the outcome
- Backlog ingestion (Linear / Jira / GitHub Issues) into Task chains
- Multi-repo migration playbooks delivered as a service
- Recurring dependency hygiene (CVE patches, library upgrades on cadence)
- Monthly outcome reports — throughput, rework, time-to-deploy
- SSO / SAML / SOC 2 / audit log export
- 24/7 incident response
Every plan includes
Custom domains with HTTPS · Application and build logs · Slack notifications · Automatic deployments on git push · Automatic previews on pull requests · Community support · Unlimited static sites, web services, repositories, and environments
All plans run in your own cloud — your code, credentials, and data stay in your environment.
Guardrails
Safe by default
Every Task runs with hard limits and human gates — sandboxing, runaway detection, approval flows, and a full audit trail. On every plan.
Sandboxed execution
Every Task runs in an isolated Docker container — non-root, read-only filesystem, dropped capabilities. The agent can't escape its work directory.
Hard runaway limits
Per-Task turn cap, token budget, and stall detection. The orchestrator kills runaway agents in real time.
Production approval gates
Production deploys pause for human sign-off. RBAC scopes who can approve which environment.
Full audit trail
Every agent action — file edits, tool calls, deploys — logged to your cloud for compliance review.
Managed playbooks
What we ship under Managed
Three patterns we run end-to-end — each change verified in a preview, then deployed to your cloud.
Multi-repo migrations
“Upgrade Stripe SDK from v8 to v17 across 5 services.”
We open dependent Task chains — the library lands first, consumers follow. Each change is verified in its own preview, then deployed in order.
Recurring dependency hygiene
“Keep our deps fresh and our CVEs patched.”
Monthly Tasks find CVE-flagged libraries and ship patched changes, each verified in a preview. Weekly digest of what shipped.
Backlog → shipped
“Drain our Linear board.”
We pull from your tracker, run Tasks against the queue, and deploy the changes — each verified in a preview. Monthly outcome report follows.
Frequently asked questions
What's the difference between Self-serve and Managed?
Self-serve (Free / Team) is the platform — describe an outcome and our agents ship it, or point your own coding agent at it over MCP. You operate it.
Managed adds our team. We onboard your codebase, ingest your backlog, run the platform, and ship the outcome to your cloud. We own throughput and verification, and pick up on-call when something fails. Both run in your own cloud — the difference is whether you operate it or we provide the engineering.
What's actually included in Managed?
A named engineer who learns your codebase, sets up dependent Task chains, and tunes the agents for your conventions. Weekly intake from your tracker — each Task verified in a preview, then shipped to your cloud.
Plus committed throughput, monthly outcome reports, SSO/SAML, audit log export, and 24/7 incident response.
Pricing is a custom annual retainer. Book a discovery call to scope an engagement.
How do you make sure changes work before they ship?
Every Task runs through a self-verification step. Our agents stand up a preview environment with the changed services (and dependent ones, like a database), run checks against it, and only mark the Task done when verification passes.
Nothing ships until verification passes. On Self-serve, the preview URL comes with the pull request — click and see, not just read diffs. On Managed, our on-call team catches anything that fails before it reaches you.
Can you handle multi-repo migrations?
Yes — that's a core Managed workflow. Tasks declare dependencies on each other, so we coordinate the changes across multiple repos, verify each, and deploy them in the right order.
The library upgrade ships first; the services that consume it follow once it's live. We own the chain end-to-end and surface progress in your weekly digest.
What counts as a job?
A Task is 1 job. So is a deploy, a log fetch, a custom domain setup, a response-header change, or a deployment deletion. A preview environment counts 1 job per service it spins up.
Multi-Step Tasks count each Step as a job, but most Tasks default to zero user-defined Steps — so they're 1 job.
Can I use cloud credits?
Yes — your cloud credits fully cover your infrastructure costs on Self-serve and Managed alike. our guide on how to get free cloud credits.
Do you have special discounts for startups?
Yes — please email us at hello@deployment.io