AI编码面临QA瓶颈，Momentic发布测试平台

AI coding has a QA problem.

And Momentic just released a testing platform around that bottleneck: faster code means little if QA cannot scale with it.

So Momentic's testing agents read the same product memory humans use, including Linear tickets, Notion PRDs, and PR diffs, so tests are based on intended behavior rather than blind script matching.

They are claiming in the past few weeks, its agents analyzed 70K+ test failures, created 600 tests, and reached a 73% PR merge rate.