Launching SentinelQA: AI-powered test insights for debugging CI failures

Hey IH folks, I’ve been building a tool to make debugging CI failures way less painful and just launched it today.

It’s called SentinelQA, an AI layer that explains failing tests, detects flaky behavior and highlights regressions without having to scroll through thousands of log lines.

The idea came from dealing with pipelines that failed for unclear reasons, tests timing out, intermittent failures, subtle regressions or logs so big they’re impossible to parse. I realized I was spending more time figuring out what happened*than actually fixing it.

So I built something that:

• summarizes each run in plain English
• detects flaky tests using historical patterns
• identifies what changed since the last good run
• surfaces action items and possible root causes
• supports visual diffs and code-level insights

Launched on HN + Twitter today and now sharing here. I would love honest feedback from other builders, especially anyone working with automated tests or CI.

Here’s the website if you want to check it out: https://sentinelqa.com

If anyone else here has launched a dev tool recently, I’d love to hear how you approached distribution. This part is kind harder than building 😅