# Sonarly

> Sonarly is a San Francisco startup (Y Combinator W26) building an autonomous AI on-call engineer for production software. It wires into a company's monitoring stack - Sentry, Datadog, Grafana, Slack and more - triages every alert to strip out noise and duplicates, investigates logs, traces, metrics and code to find the root cause, and opens a ready-to-merge fix pull request before a human engineer is paged. Founded by Dimittri Choudhury and Alexandre Klobb, who previously scaled an edtech app to 100k+ users, Sonarly aims to cut Mean Time To Repair by handling the 3am pages that engineers dread.

- **Founded:** 2025
- **Headquarters:** San Francisco, United States
- **Founders:** Dimittri Choudhury (Co-founder & CEO), Alexandre Klobb (Co-founder & CTO)
- **Team size:** 2
- **Products:** Sonarly AI on-call engineer, Living production map, Alert triage & deduplication
- **Notable:** Accepted into Y Combinator's Winter 2026 batch, Reported 78% root-cause-analysis accuracy versus 53% for a raw agent with MCP monitoring connections, Featured among TechCrunch's most interesting startups from YC W26 Demo Day

## Products & services

- **Sonarly AI on-call engineer** — An autonomous agent that connects to a company's monitoring stack (Sentry, Datadog, Grafana, Slack, Discord, Linear, GitHub, PagerDuty), triages and deduplicates alerts, investigates root cause across logs, traces, metrics and code, and opens a ready-to-merge fix pull request or recommends alerting-rule changes.
- **Living production map** — A dynamically generated markdown map of a customer's production system - services, dependencies, logs and historical alert patterns - that gives the coding agent enough context to raise root-cause accuracy from 53% to 78%.
- **Alert triage & deduplication** — Clusters duplicate alerts and filters false positives, reducing a flood (e.g. ~180 alerts from a bad deploy) down to a handful of actionable issues before anyone is paged.

## Achievements

- Accepted into Y Combinator's Winter 2026 batch
- Reported 78% root-cause-analysis accuracy versus 53% for a raw agent with MCP monitoring connections
- Featured among TechCrunch's most interesting startups from YC W26 Demo Day
- Founders previously scaled edtech app Meoria to 100,000+ users in about 6 months

## Latest updates

- **2026-03** — Publicly launched via Launch YC and Launch HN as 'the AI engineer for production' and appeared at YC W26 Demo Day.
- **2026-02** — Listed on Product Hunt as 'The AI that fixes prod autonomously,' powered by Claude Code and Opus 4.6.

## Links

- Website: https://sonarly.com
- LinkedIn: https://www.linkedin.com/company/sonarly-dev
- Twitter/X: https://x.com/dachoudhury

---

Profile page: https://yespress.io/sonarly-yc-w26
Published by YesPress — https://yespress.io
Last updated: 2026-07-30