Rootly

Nothing bonds engineers more than AI… or a good World Cup match. Rootly hosted a happy hour with @Tailscale, @AikidoSecurity, @Docker, and @inngest to help folks in San Francisco for the AI Engineer World Fair wind down. Thanks to the 400+ engineers who stopped by to share your passion for AI and soccer with us.

The AI World's Fair took SF by storm last week. Rootly teamed up with @mezmodata, @TwingateHQ, Echo, and @RespanAI to host an after-party with award-winning craft beers and engineers pushing the boundaries of AI. Thanks to the 150+ builders who stopped by!

Learn more: root.ly/ai-retro-templ…

Rootly is introducing a new way to draft AI retrospectives tailored exactly to your needs: customizable blocks with precise sources and instructions. Each block streams data from your incident data, Slack channels, and call transcripts. 6 block types come by default: summary, impact, root cause, curated timeline, mitigation, and resolution. You have complete insight into how every section of the retrospective came to be. You can define your own block types with actions tied to them and use them to build retrospective templates. Have your retrospective write itself, exactly the way you want it. Review it, refine it, and publish it confidently.

Read more about Ankita's view on reliability: root.ly/ankita-sre

Ankita Gandhi now drives reliability at @glean. But… when she was in college, she didn't think it'd be her thing. Even if not fully convinced, when an internship at the company her sister worked at came up, she took it, as one does in college. Fast-forward a few years: Ankita has worked at Apple and Goldman Sachs as a site reliability engineer, always with a focus on proactive reliability. For Ankita, each incident is an invitation to learn. But improving the system to prevent them is where the real work lies. "My team gets involved early in the release process, before anything goes out. What could we miss? What won't work in terms of observability?" says Ankita on how Glean builds reliability as a feature early in each release.

Adib Fallahpour, researcher at @nvidia, shared how he’s using AI during our full-room Codex event in Toronto. The evening was full of in-depth talks that helped us imagine new ways of using AI agents beyond coding, including biology. Thanks to Muhammad Hamza, @OpenAI Ambassador, for co-hosting, and to the 400+ who signed up. The room only fits 150, so we're already planning the next one.

Full episode + notes: root.ly/stop-shipping-…

Not a lot of leaders are bold enough to stop feature development for an entire quarter. But for @ekampf, VP of Engineering at @TwingateHQ and co-founder of @mondaydotcom, it was crystal clear: the team had to focus on reliability first. Because incidents don't just cost engineering time. They cost trust. Every incident sends support into damage control, executives onto apology tours, and engineering loses focus. Rebuilding trust meant rebuilding the system: – Single cluster to active-active multi-region. – Sharded production so a bad deploy only hits a fraction of customers. – And a clear model for where AI belongs in incident response, and where a human has to stay in the loop. Eran and @SylvainKalache, Head of Rootly AI Labs, on making reliability a feature, not a byproduct.

Stack Overflow has answered millions of developer questions. Rootly answers when their platform needs it most.

Last week, we wrapped up PlatformCon in NYC with hundreds of platform engineers and our friends from @ClickHouseDB, @TwingateHQ, and @tweetsbyport. Even though the World Cup was playing on the screens, everyone kept chatting about their favorite talks of the day and how they’re enabling their org with AI through their platform. Thanks everyone for joining us!

Want to work at @nvidia? Knowing @rootlyhq now gives you an edge. NVIDIA just posted a Senior Resilience Engineer role for DGX Cloud — the infrastructure the AI industry runs on. The listing has a section called "ways to stand out from the crowd" — the things that make a reliability engineer exceptional. One bullet: experience from a world-class reliability function like @googlesre or Meta production engineering. Another: proficiency with Rootly. Google SRE pedigree and knowing Rootly, side by side, both filed under what makes a great reliability engineer. It's been incredible to watch how fast Rootly has become the standard for the best engineering teams at every scale. Something I am very proud of. And the best is yet to come! P.S. — the NVIDIA team is truly next level and a joy to work with. Link to job posting in thread.

@ahmednadar What Iain presented is the latest version, with significant more AI capabilities. But earlier this year one of our engineers wrote about how we built the editor iteration at that time, might have something useful? - Jorge L rootly.com/blog/how-we-us…

The combination of real-time collaboration and AI-native support in a single Rails app is the architecture question I've been working through. ActionCable handles the sync layer well, but feeding inference results back in real-time is where things get interesting. Would love to see Ian's approach written up somewhere.

We welcomed the Ruby Toronto community at Rootly HQ for two deep dives: – Ian (Rootly) on the intricacies of building a real-time collaboration editor that supports AI natively – Andrew Novoselac (Shopify) on how to use mutable data structures that are Reactor-safe in Rails Thanks everyone for coming out! It was great to hear how you're all building with Ruby.

Kickstarter lets millions of backers fund bold ideas. Rootly keeps the Kickstarter platform reliable for all of them.

The expo floor closes. Time for the AI conversations you won't find on stage. If you're at AI Engineer World's Fair, join Docker, @Tailscale, @AikidoSecurity, @inngest, and @rootlyhq for drinks, darts, and good company. 📍 Golden Eye Social, SF 🕕 July 1, 6PM Register → lu.ma/2avil0ni

Learn more: root.ly/cortex

Your service catalog says one thing. Your org chart says another. Incidents are where you find out which one was right. When your catalog lives in one tool and your incident response runs in another. Who owns what drifts... and you only notice at the worst possible moment. Rootly now syncs directly with Cortex, to prevent any of that. Cortex stays the source of truth for entities in the catalog, and Rootly keeps all references updated. Plus, you can add Rootly-only fields to the entities. The point isn't tidy catalogs. The point is that when something breaks, the responder is working from how your systems are structured today, not how they looked last quarter.

We gathered some of the best engineering leaders in Manhattan after AWS Summit NYC in a speakeasy and let them figure it out. No AI slop, but plenty of chats on how new infra and reliability challenges keep popping up as teams find themselves having to scale LLM products, deal with unpredictable outputs (and costs) of frontier models, and shift an engineering culture away from assumptions that are no longer valid. Thanks to our partners @cloudsmith, @Mend_io, @GetCortexApp, @ClickHouseDB, and @Docker for making it a great evening.

.@runwayco helps finance teams forecast what's next. One thing Runway doesn't have to forecast: uptime. Rootly's got it for them.