Rootly

Responders can now get paged directly via Slack DM. When an alert fires, you’ll receive a direct message from the Rootly bot with full context so you can acknowledge, resolve, or escalate instantly, without switching tools. What this unlocks: • Faster acknowledgment with fewer clicks • Lower cognitive load during high-stress moments • A consistent responder experience across web and Slack • Less context switching, especially on mobile or during off-hours Learn more in our changelog: https://lnkd.in/e9yA2uU4

Growth at this stage is only as good as the people behind it. Meet our latest. → Marcela Havandjian Begalli, Principal Product Designer → Marshall C., Staff Product Designer → Sébastien Gruhier, Member of the Technical Staff → Harneet Singh, Member of the Technical Staff → Andrew Parnell, Senior SRE When engineers and designers at this level choose to join, they're not betting on potential: they're betting on momentum. NVIDIA. Canva. Figma. Wix. Hundreds of the world's best engineering teams trust Rootly with their most critical workflows. This team is here to make sure we keep earning that. Welcome to Rootly. 🙌

We stopped reading candidates' code. We read their AI transcripts. For years, our take-home at Rootly was code-first: edge cases, anti-patterns, structure, naming. The code told us how someone thinks. That was enough. But, it's not enough anymore. When everyone has access to Claude Code, Cursor, or Copilot, clean well-structured code is table stakes. Most companies have responded by banning AI in interviews or pretending it doesn't exist. Both are wrong. AI is how engineers work now. The question was never whether candidates use it, it's whether they use it well. So we made AI session transcripts a first-class submission artifact. Not an optional add-on. We recently reviewed two submissions with nearly identical code quality: clean Rails, solid test coverage, working Slack integration. The transcripts told completely different stories. One candidate broke the problem into clear phases, asked targeted questions about Slack's API edge cases, and course-corrected when the generated code didn't handle webhook retries. The other pasted the full spec into a single prompt and re-prompted "fix it" until it worked. Same output. Completely different engineer. Boris Cherny, creator of Claude Code, described it well in a recent YC Light Cone episode: "You can figure out how someone thinks… when they use plan mode, do they make sure that there are tests… do they think about systems? There's just so much embedded in that." I couldn't have said it better. Reading an AI transcript is watching someone think out loud. The behavioral dimension: how someone decomposes a problem, handles failure, iterates, has always mattered in engineering. AI just made it directly observable. My full write-up if you're rethinking your own process: https://lnkd.in/gYyQmr3b

Join Sylvain Kalache (Rootly), Philippe Ozil (Salesforce), and Ming Z. (AWS) as they break down how they’re using Postman to drive real API adoption. In this virtual session by Postman, they’ll cover: – API adoption strategies that actually move usage – Hard lessons learned scaling API programs – Enablement best practices for builders – How they’re leveraging the MCP Catalog RSVP here: https://lnkd.in/eERbPwQm

Rootly’s Collaborative Retrospectives run on real-time sync, deep workflow engine integrations, and built-in AI assistance. To make that seamless, we leaned on Conflict-free Replicated Data Types (CRDTs). Michael Han breaks down how our engineering team built this feature, from implementing hydrated liquid variables to shipping three custom dynamic editor extensions that accommodate live incident data. Read the full breakdown: https://lnkd.in/e-itJwiA

Large incidents rarely move in a straight line. You have engineering driving root cause, customer teams managing comms, security running parallel reviews, and sometimes regional teams coordinating locally. Without structure, that turns into fragmented updates and duplicated work. With Rootly’s Sub-Incidents and our new workflow action, you can treat the parent incident as the single source of truth while sub-incidents execute in parallel and stay automatically in sync. When the Create Sub-Incident action runs: • Centralized truth, distributed execution • Automatic status alignment across work streams • Intentional stakeholder separation • Aggregated visibility for leadership • Synchronized Slack communication Parallel response without losing control. Learn more in our changelog: https://lnkd.in/eBqnJUMU

Throughout his career, Sylvain Kalache has watched friends and colleagues burn out while being on-call. The hardest part? Responder exhaustion is difficult to measure and quickly becomes merely anecdotal. That’s why we built On-Call Health, an open-source project that connects to the tools your responders already use: Rootly, PagerDuty, Jira, Linear, GitHub, Slack. It benchmarks workload for teams and individuals against their own baseline. Because on-call is personal. What feels routine for a 10-year veteran can be overwhelming for someone just getting started. SREs measure the health of production systems by default. It’s time we measure the health of the engineers running them. Try it today: https://lnkd.in/eW4fuPEB

The idea that “this time we won’t need engineers” isn’t new. We’ve been here before. Swizec Teller, bestselling author and speaker, and Sylvain Kalache, Head of Rootly AI Labs, zoom out from the AI hype and look at the bigger pattern: Tools get easier. Prototypes get faster. But when systems run in production, ownership always comes back to engineers. If code is becoming cheap and probabilistic, reliability becomes the real skill. Production readiness. Clear ownership. Real accountability. Listen to the full episode, Code Is Cheap, Reliability Isn’t: https://lnkd.in/eC57hYbV