Write it down, replay it, ship it: why we build ERDs to build product
How we documented Rootly's proactive incident agent as an interactive ERD you can click, replay and check against the code, and why a static design doc couldn't carry it.
Engineering Blog
Investigating CVE-2026-66066 with the Rails Forensics Agent Skills
How I used Claude to investigate KindaRails2Shell and sleep better after the patch.
We pay engineers to cut our infra bill
Up to $25k for finding and shipping cost savings initiatives.
Why we got rid of our small-PR rule
Small PRs are great for humans, but not for AIs. Instead of focusing on diff size, we moved ot a risk-based model.
What broke when engineering went fully agent-based
Fragmentation compounds really fast when everyone ships with their own agenting workflow
AI changed the QA job, the skepticism stayed.
In a typical week, our team ships four or five features. Sometimes more. Testing all of them at that pace, with humans, isn't really realistic.
Stop trying to review AI's code faster: bet on rollback instead
AI's failure mode has changed. Most incidents aren't code bugs: they're context bugs. And faster review can't catch them. A case for rollback.
Replay, don't rebuild: adding week-long deferrals to a pipeline that handles millions of alerts
On extending a high-volume production pipeline instead of building alongside it, and the edge cases that only appear at the intersection of correct decisions
A council, a sword, and a fleet of agents: how I ship code now
Instead of writing code, lot of the work I do is providing context to the agents so I can, hopefully, come back to them having completed a bunch of tickets.
How we used CRDTs to build Real-Time Collaborative Retrospectives
A dive into how we built collaborative software with real-time features, AI assistance, and workflow automation with Y.js CRDTs).
Alerting as Code: How Mistral AI Uses Terraform as the Source of Truth
A Terraform-first model for deterministic alerting in AI systems
Prototyping with design playgrounds
Moving design decisions from opinions to action. Real products are dynamic, interactive, and a little subjective. That’s why design decisions over a static Figma file are not always be easy. At Rootly, we build small, standalone design playgrounds to feel the work sooner and make better decisions fa
How we built an OSS LLM-powered Incident Diagram Generator
Discover IncidentDiagram, an open-source CLI tool that uses LLMs to turn incident retrospectives and codebases into easy-to-understand visual diagrams.
Page 1 of 2 · 16 articles