# We want CI to stop being a thing you babysit.

Source: https://latchkey.dev/about

Latchkey is building the autonomous CI/CD layer for the agent era: build pipelines that detect, diagnose, and repair their own failures instead of waiting for a human to read the logs.

## Our mission

Our mission is to remove the human from the loop of maintaining software delivery. Not the judgement calls, and not the decisions about money and risk, but the mechanical work: reading a failed build, finding the one line that broke it, applying the fix, and running it again.

AI now writes code faster than teams can maintain it. The volume of change went up, the pipelines did not get smarter, and CI became the bottleneck. So Latchkey sits inside the build step rather than reporting on it afterwards: when a command fails we already have its full output, and we can repair it while the job is still running, then open a pull request so the fix stays fixed.

Anything we cannot fix leaves as a structured failure report that a coding agent can act on directly. The goal is a pipeline that answers for itself, so the only thing left for a person to decide is what is genuinely worth deciding.

## Elsewhere

### Careers

Open roles and how we work.

### For agents

The API, the CLI, and the MCP server your agent can call.

### Contact us

Questions, feedback, or a hand getting started.

## The team

Two founders, both from the infrastructure side of shipping software.

### Daniel Zoghalchali

CEO & Co-Founder

Previously in the gaming industry, working on developer infrastructure and security at very large scale.

### Kaveh Alemi

CTO & Co-Founder

Previously in cloud, building secure infrastructure for enterprise and public-sector teams.

---

Latchkey runs CI/CD that repairs its own failures. Agent entry points: https://latchkey.dev/agent.txt, https://latchkey.dev/openapi.json, https://latchkey.dev/llms.txt