Humans of Reliability

Your AI agents are lost: give them a graph

Your AI agents are lost: give them a graph

Anthony Alcaraz

GTM Agentic Engineering Lead at AWS

Bigger models cannot fix missing context. The real advantage may belong to teams that teach agents where to look, what to remember, and when to learn.

Podcast
Your next internal developer platform is a library of agent skills
Michael Kutsch
Staff SRE and Team Lead, Cloud Foundations at PostHog
PostHog's Michael Kutsch explains why agent skills are becoming an internal developer platform, how autonomy drives adoption, and where deterministic guardrails still matter.
Blog
Nobody wants to review 1,000 lines of AI-generated code
Engineering leaders from CircleCI, Airbyte, Flare, Moloco, and Twingate reveal what their teams use instead of traditional code review.
Sylvain Kalache
Podcast
Getting hired as an SRE is changing
James Duffy
Principal SRE at Fabric
Fabric Principal SRE James Duffy explains how AI is rewriting the SRE career path, from getting hired and developing judgment to continuous learning and AI platform engineering.
Podcast
AI vs. AI: From Alert Fatigue to Agentic Cybersecurity
Nir Soudry
Head of R&D at 7AI
Attackers don't need sleep—and now their malware can make decisions. Nir Soudry explains why security's next race is AI against AI, and where human accountability still has to win.
Podcast
We Stopped Shipping Features for a Quarter — Here's Why It Saved Us
Eran Kampf
VP of Engineering at Twingate
What happens when reliability debt starts consuming the whole company? Eran Kampf stopped feature work for a quarter. The gamble—and the architecture that followed—changed how Twingate ships.
Blog
Borrowed gravity: words worth changing
Some words borrow a gravity they haven't earned, others shrink the job to its smallest piece. On "firefighting," "war room," "postmortem," and "AI SRE".
Adam Frank and Sebastian Vietz
Profile
Ankita Gandhi never stops at fixing the incident
Podcast
"Hey Claude, where's my database?": How an AI agent nuked production
Alexey Grigorev
Founder of DataTalks.Club
Claude Code warned him. Alexey Grigorev overruled it. Minutes later, his production database was gone. Here's the chain of decisions behind the outage—and the guardrails he trusts now.
Podcast
Every pilot is ready for engine failure: are your engineers?
Hamed Silatani
Co-Founder and CEO at Uptime Labs
Pilots rehearse engine failure before it happens. Most engineers meet their first major incident live, in production. Hamed Silatani wants to change that—and argues AI makes training more urgent.
Blog
Re-thinking candidate take-homes in the AI Era: transcripts over code
When everyone can ship clean code with AI, the transcript of how they got there becomes a key signal.
Quentin Rousseau
Podcast
LLM Observability: Lessons From MLOps
Maria Vechtomova
Co-founder, Cauchy
MLOps spent years warning teams to version, monitor, and roll back. Then LLMs made everyone listen. Maria Vechtomova explains what truly changed—and what teams still refuse to learn.
Podcast
The golden path to nowhere: When platforms undermine reliability
Chase Roberts
COO at Northflank
Chase Roberts, COO at Northflank, challenges the conventional wisdom around internal developer platforms (IDPs), making a compelling case that complexity—not just code—may be your team’s biggest reliability risk.

Page 1 of 9 · 99 entries