The people, decisions, and hard-earned lessons behind reliable systems.
Podcast
Fabric Principal SRE James Duffy on rebuilding his SRE interview around AI-generated Terraform, why junior hiring still matters, and continuous learning.
“Your AI is not the risk. Your blind trust in it is.”
Dear Ben On-Call,
I’ve always been the incident commander who knows which dashboard to open and who to page.
Now AI is doing more of the triage and writing the timeline. If the tools keep taking over the things I’m good at…
What makes me a good incident commander?
—Formerly the Fastest
Ask Ben On-Callben@rootly.com
Ben McBratneyHead of Incident Strategy at Rootly
Dear Formerly the Fastest,
As incident commander, knowing where to look and who to call has served you well. But the most valuable part of that experience is knowing what to do with what you find.
As AI takes on more of the correlation, first-pass triage, and timeline drafting, your job shifts toward judgment.
Those decisions require someone who understands the system, the business, and the consequences of being wrong. Your experience gives you a basis for questioning a recommendation, including one that arrives quickly and sounds confident.
Expect your tooling to tell you what’s happening, why it matters, and what evidence supports its recommendation. You need enough context to make a decision—and to recognize when you need more.
That’s the role we’ve designed Rootly AI SRE to support: an incident sidekick that helps you spend more time on the decisions that need you.
Being first to open the right dashboard was useful. Knowing whether to trust what’s on it still is.
—Ben On-Call
Profile
The Glean SRE treats every alert as an invitation to improve the product, a habit that has shaped eight years of reliability work.
“What did I do to make the product better? That's the measure I like to follow.”
Interview
Engineering leaders from CircleCI, Airbyte, Flare, Moloco, and Twingate reveal what their teams use instead of traditional code review.
Podcast
7AI's Nir Soudry on AI-driven attacks, onboarding agentic security like a junior analyst, accountability when AI investigates incidents, and less alert noise.
Podcast
Twingate VP of Engineering Eran Kampf on pausing features for a reliability quarter, the multi-region architecture that followed, and AI in critical systems.
Podcast
DataTalks.Club founder Alexey Grigorev on how an AI agent deleted his production database, how he recovered, and the guardrails he uses now.
Podcast
Cauchy co-founder Maria Vechtomova on applying MLOps lessons to LLM observability, versioning and rollbacks, business metrics, and skills SREs need for MLOps.
Podcast
Northflank COO Chase Roberts challenges conventional thinking on internal developer platforms and explains why platform complexity can be a reliability risk.
Podcast
Cliff Snyder of Multimedia LLC on scaling incident response from 600 SREs to 6,000 engineers over 18 months, and why engineer-hours revealed more than MTTR.
Podcast
Cortex CTO Ganesh Datta on service ownership, platform engineering as a product, production readiness standards, and how AI exposes engineering bottlenecks.
Podcast
Antithesis's Will Wilson on deterministic simulation, lessons from FoundationDB, reproducible fault injection, and why AI-written code raises the stakes.
Podcast
Stephen Townshend, SRE team lead at IAG, on his burnout and year-long recovery, the warning signs he missed, and how autonomy, slack time and org design help.
Podcast
Author Swizec Teller on why users buy reliable services rather than code, managing AI coding agents, writing useful logs, AI SRE benchmarks, and accountability.
Article
Introducing On-Call Health, a free open-source tool that connects to Rootly, PagerDuty, Slack and GitHub to detect responder overload before it becomes burnout.
Stories couldn’t be loaded.
No stories match these filters. Try another topic or change your search.
