SAN FRANCISCO, CA — Momentic has launched a significant platform replace that rethinks software program verification for the AI period. AI coding instruments have made it doable for groups of all sizes to ship extra code than ever earlier than, however how that code will get validated has not saved up. The result’s bugs reaching manufacturing sooner than groups can catch them, and engineering groups caught firefighting hotfixes as a substitute of delivery.
Each Momentic agent — whether or not constructing new checks, classifying failures, or therapeutic damaged ones — runs on a shared information base of how your product really works. Groups can outline terminology, important flows, and customary failure patterns; Momentic maps each consumer journey and retains that understanding present because the product evolves. It ingests Jira tickets, GitHub, help docs, Slack, and connects with different inner instruments, so the system at all times is aware of what’s altering, what’s breaking, and whether or not a failure is an actual bug or from an supposed change. The result’s a system that will get smarter the extra you utilize it.
Key capabilities on this launch embody:
Check Protection That Grows With Your Product
Most groups share the identical drawback: protection gaps that widen each dash as a result of writing new checks doesn’t scale with the velocity of delivery. Momentic’s Discover Agent closes that hole mechanically. It screens pull requests, reads diffs and codebase context, identifies which flows lack protection, and proposes new checks or updates present ones, with none guide authoring.
Failure Classification and Therapeutic: From Noise to Sign
Flaky checks and noisy CI pipelines are one of many largest drains on engineering time, and one of many important causes groups cease trusting their check suites. Momentic’s Failure Classification Agent mechanically categorizes each failure: actual bug, supposed utility change, check setup challenge, or transient error. When a failure isn’t an actual bug, Momentic proposes a repair and opens a PR. When it’s, the staff will get a high-signal alert with full context on precisely what broke and why.
Legacy check frameworks produce artifacts that solely the engineer who wrote them can perceive. Momentic’s new check format is intent-based and readable by each people and AI brokers. Engineers describe what they wish to check in plain English; Momentic executes it stay towards a browser or emulator.. AI brokers can parse, construct, and modify checks extra successfully, making your entire improvement loop sooner and extra autonomous.
Momentic’s clients are already seeing the advantages of its launch with measurable affect. In just a few weeks of beta entry alone, Momentic’s brokers analyzed over 70,000 check failures, opened 400 pull requests, and wrote 600 checks with a 73% merge price. In complete, clients have executed greater than 2 billion steps on the platform and verified over 80,000 PRs: proof that agentic verification doesn’t simply sustain with the AI coding instruments, it really offers engineers the boldness to ship extra with them, sooner.
For extra details about Momentic’s availability and pricing, go to momentic.ai/pricing.
