Failed Builds on Main, Triaged Automatically
When CI fails, an agent reads the failure, finds the likely cause, and opens a fix when the cause is clear. When it isn't, it reports what it found.
A Red Build Blocks Everyone, and Nobody Owns It
A failed build on main stops merges for the whole team, but the fix often waits until someone notices, reads the log, and works out which change broke it. Most of that time goes to the same steps: find the failing job, read the error, compare it with recent commits.
An Agent Reads the Failure and Proposes the Fix
Your CI system sends a failure event to Felan, and an automation starts an agent with the event context. It runs the relevant checks in an isolated environment, finds the likely cause, and opens a pull request with the fix when the cause is clear. When the cause is uncertain, it reports its findings instead of guessing. A person reviews the fix before it merges.
test/users/serialize.test.ts
✕ TypeError: user.profile is undefined
src/users/serialize.ts
- id: user.profile.id,
+ id: user.profile?.id ?? user.id,
Fix opened, checks green
What Teams Report
9.4x
growth in weekly agent requests, February to August 2026
Uber runs managed agents that self-heal CI failures, with human review or escalation, alongside code review and alert triage.
Uber's software factory postCommon Questions
How does Felan know a build failed?
Your CI system sends an event to Felan's team webhook endpoint, and an active automation that matches the event starts the agent.
Does the agent merge its own fix?
No. It opens a pull request for a person to review. Merging stays a human decision.
What if the cause isn't clear?
It reports what it found, including the failing step and likely suspects, instead of guessing at a fix.
Which models can it use?
Any major provider, plus local and private models. Bring your own keys or subscription, and there's no markup on tokens.
Can it run on our own infrastructure?
Yes. Use Felan's managed cloud, or run Felan on your own infrastructure with the Enterprise plan.
Related: Webhook events docs, Flaky test cleanup, First-pass code review.
Keep Your Attention for the Hard Problems
Start with one repo and one recurring task.