IntakeAnyone files from Claude or ChatGPT, in a real conversation.The board and DocsEvery request with its brief, size, worth and seat, and the specs and decisions beside it.Smart routingThe best seat and the best model for every request.The recordEstimate against actual, signed. Better with every row.ConnectionsClaude, ChatGPT, Claude Code, Codex, GitHub, Slack and more.
Business teamsSales, success, ops: send the customer's request and hear when it lands.Product teamsSpecs and prototypes that arrive ready to build, and a say in what runs.EngineeringThe right work, on the right seat and model. For engineers and leads.
Docs and helpEvery verb, every surface.Developerscurl -fsSL sleds.ai/install | shTrustWhat we keep, and the exceptions.ManifestoWhat we are doing, and why.Claude connectormcp.sleds.ai/mcpSelf-hostThe record in your cloud, on your keys.ContactTalk to the people building it.
Request an inviteSign in
sleds/Company/Careers
careers · a small team in San Francisco

Build the thing the agents run on.

sleds is where the business meets engineering. We are a small team, and most of what we ship is shipped by agents we orchestrate ourselves. The people we hire build the system underneath that: the intake, the engine, the record. One role is open.

EngineeringSan Francisco · hybrid, 10–20% in the officeFull-time

Software Engineer, Developer Productivity

Intake from Claude and Slack, work sized as a band, routed to a seat, run by an engine on the engineer's own machine, and landed with a receipt. The infrastructure those agents run on is this job.

What you will own

  • The CI every route lands through: test selection, flake detection and quarantine, merge queues across a monorepo, builds that scale with hundreds of agent sessions a day.
  • The engine's runners: tmux, headless and cloud seats that start, follow and recover sessions, on Macs and in containers, without a person watching.
  • Observability and incident automation for a system whose users are agents: load shedding, automatic recovery, and the record of what ran and what it cost.
  • CLI tools and the GitHub App that carry the brief in and the receipt out.

You might be a fit if you

  • Have built and operated CI/CD at scale (Buildkite, GitHub Actions, Jenkins or similar) and have opinions about merge queues and flaky tests.
  • Have run container orchestration in anger, on more than one cloud.
  • Care about developer productivity as a product: removing friction, measuring it, never failing the same way twice.
  • Have used coding agents for real work and want to build the system that runs them.

Nice to have

  • Intelligent test selection; GitHub API automation; macOS tooling; Rust or TypeScript.
  • Having filed a bug report that an agent fixed before you finished your coffee.

How to apply

Write to us through the contact page with the role in the first line, a link to something you built or ran, and a few sentences on the worst CI incident you have been inside and what changed after it. We read everything ourselves and reply to everyone. We encourage you to apply even if you do not meet every line above.

How we work. Every request we build is sized before it runs and closed with a measured actual, including the ones this role will own. Read the manifesto and news for the rest.
not the right role

Tell us what you would build.

If you think sleds needs a role that is not listed, say what it is and why now. That is a request, and we size those too.

Contact usOr request an invite and build with it first.