February 17, 2026
Paging Dr. Robot
Launch HN: Sonarly (YC W26) – AI agent to triage and fix your production alerts
AI to fix your app’s 3 a.m. meltdowns? Devs cheer—then groan at a sluggish signup
TLDR: Sonarly claims an AI can read error reports and open code fixes to cut downtime, but an early tester hit a slow, confusing signup. Commenters split between dreaming of fewer late‑night alerts and worrying about a robot tweaking alerts and shipping PRs without enough oversight—relief versus risk.
YC newcomer Sonarly waltzed onto the scene promising “from alert to fix” in four steps and no more soul-crushing triage. Plug it into Sentry, Datadog, or your tools, it says, and the AI will group duplicate errors, silence false alarms, rank what actually hurts users, even tweak your alert settings to kill the noise. Then it hooks GitHub to open a fix as a pull request. Short version: fewer 3 a.m. wake-ups, faster “mean time to repair” (the clock from break to fix).
But the comments lit up when one early tester, jefflinwood, hit the brakes: onboarding hung on “Analyzing” when there were no Sentry issues; it eventually let him in, but after a head-scratching wait. That single hiccup became the day’s meme: “The robot that fixes everything can’t get past step one.” Some devs cheered the thought of the “on-call intern that never sleeps.” Others side‑eyed the idea of a bot changing alert settings and rushing out code fixes—cue debates about trust, reviews, and “what if it hides a real fire?”
The hottest split: folks who’ve done the “one-person-per-sprint” bug duty dreaming of winning their weekends back, versus skeptics asking who’s on the hook when the bot ships a Friday PR. Jokes flew about “PagerDuty unemployment,” “AI ate my wake-up,” and the classic: “Great… until it closes my alert, not my outage.”
Key Points
- •Sonarly automates triage and fixes for production alerts to reduce MTTR.
- •It ingests errors from observability tools like Sentry and Datadog with full context.
- •AI clusters duplicates, filters false positives, and ranks alerts by real user impact.
- •The system updates alert configurations to reduce noise and make alerts actionable.
- •By integrating with GitHub, the agent investigates code and opens pull requests with fixes.