Operations | Monitoring | ITSM | DevOps | Cloud

Un-observable AI is Un-trustworthy AI

Recently, someone talked Chipotle’s customer support agent into reversing a linked list – a task completely unrelated to burritos in any way. Screenshots circulated, people laughed, but underneath the joke sat a sharper question. If a production support agent will do that on a public channel, what else will it do that nobody is screenshotting? The bug is funny. The trust gap behind it is not.

Measuring engineering organizations in the age of AI

Engineering leadership is in the middle of a real transition, and most of the leaders I talk to know it. AI has reshaped how software gets built quickly enough that the operating models many of us spent a decade refining no longer fit cleanly, and there is a great deal of serious work happening across the industry to figure out how these models should evolve. The teams I find most impressive right now are the ones treating their operating model as an open question rather than a settled one.

Beyond Mythos: responding to a new threat landscape

Canonical’s security philosophy has always been built on the premise that vulnerabilities exist and will be discovered. Our response relies on defense-in-depth architecture, rapid patch deployment, and strict adherence to Coordinated Vulnerability Disclosure (CVD). AI changes vulnerability discovery volume and speed. We have a robust vulnerability management process that is backed by rigorous compliance certifications.

AI pricing explained: what AI actually costs and how providers charge for it in 2026

AI pricing covers the cost structures and billing models providers use to charge for AI products: per-token APIs (GPT-4o at $2.50/1M input tokens), per-seat subscriptions (Copilot at $30/user/month), per-conversation billing (Agentforce at $2/conversation), and consumption-based GPU compute (H100 instances at $55.04/hour). There is no standard. The total AI cost is almost always higher than the sticker price.

The bottleneck has moved. AI is rewriting the Software Development Lifecycle

If you've read our previous piece on the 8 stages of AI engineering maturity, you know where your team sits. Turns out adopting AI is the easy part; adapting to its consequences is where most organizations struggle. For more than a decade, software organizations optimized around a single assumption: implementation capacity was scarce.

The Godfather of AI Ready Data Centers | OCOLO CEO & Founder Tony Rossabi

AI is reshaping digital infrastructure, but the biggest challenge isn't always building bigger data centers, it's finding the power to run them. In this episode of Uplink, Michael Reid sits down with Tony Rossabi, Founder & CEO of OCOLO, to discuss how AI is changing the data center industry and what it takes to deliver the next generation of infrastructure.

AI is only one of four things driving the data center boom

Tony Rossabi, aka the Godfather, has spent 30 years in this industry. Car washes to Telx to building data centers. He sat down with our CEO Michael Reid to break down what’s actually happening underneath the AI headlines, from where the real demand is coming from, to why a single megawatt of power is so hard to find, and how a team of eight is building 19 ten-megawatt facilities across two continents in 24 months.

Inside the AI Team Weekly: AI Observability workflows and Prometheus exemplars (May 19th, 2026)

The Grafana AI team (Engineers Ivana Huckova and Sonia Aguilar) share what's new in AI Observability this week: a new way to instrument and visualize agent workflows, plus a neat trick for jumping straight from a metric spike to the exact conversation that caused it using Prometheus exemplars. In this episode: We're showing parts of our team meetings to build in public in some small way and give you a sneak preview of what's to come. But not all features we show may make it to production! You've been warned. :)

Introducing Kepler | GitKraken's Agentic Development Environment (ADE)

Kepler is GitKraken's agentic development environment: mission control for running parallel coding agents at scale. Running one agent is easy. Running five of them across three repos is where things break: scattered terminals, no shared view, no idea what's done or stuck. Kepler puts every agent session on one surface so you can plan work, write code, and review what ships without losing track of anything.