Operations | Monitoring | ITSM | DevOps | Cloud

When I joined Kubex last year, the company was already well aware of the growing power of Large Language Models. As a company focused on intelligent resource optimization for Kubernetes, GPUs, and cloud infrastructure, generative AI didn’t feel like a threat so much as a natural extension of where the industry was heading. Kubex had already invested heavily in machine learning, but it was becoming clear that foundation models could unlock an entirely new class of capabilities for our customers.

History is everywhere at Dartmouth College, and while the campus is steeped in tradition, its IT infrastructure can’t afford to get stuck in the past. In an institution where world-class research and undergraduate studies intersect, technology must be fast, invisible, and – above all – reliable. That reliability was put to the test when Dartmouth’s load balancing vendor was acquired twice in five years, as Avi Networks moved to VMware and VMware moved to Broadcom.

Setting up alert thresholds in Graphite transforms raw monitoring data into actionable notifications, helping you address system issues before they escalate. Here's what you need to know.

Creating a custom dashboard is the best way to monitor metrics that matter most to your systems. Tools like MetricFire make this process straightforward by combining hosted Grafana and Graphite, eliminating the need for self-hosted solutions. Here's how you can build dashboards tailored to your needs.

You might expect an AI-SRE agent to target 100% reliable services, ones that never fail. It turns out that past a certain point, however, increasing reliability is worse for a service (and its users) rather than better! Extreme reliability comes at a non-linear cost: maximizing stability limits how fast new features can be developed, dramatically increases the operational cost, and reduces the features a team can afford to offer.

Imagine pulling up to a gas pump, inserting your credit card, and having the display on the pump say “denied”. You call your credit card company, and they say, “Oh, we don’t know, maybe it’s the merchant’s fault, or the card reader is bad…, we can look into it and get back to you in a few weeks.” Most of us would be pretty upset with that response.

We’ve all seen it happen. A DevOps initiative starts with high energy, but two years later, you’re left with a sprawl of "fragile agile" pipelines. Every team has built their own bespoke scripts, security checks are inconsistent (or non-existent), and maintaining the system feels like playing whack-a-mole. This is where the industry is shifting from simple DevOps execution to Platform Engineering.

As Large Language Models (LLMs) evolve from simple chatbots into agentic workflows, the need for a standardized way to connect them to external data and infrastructure has become critical. In a recent workshop hosted by Nir Adler, Innovation Engineer at Komodor, we explored how to bridge this gap using the Model Context Protocol (MCP).

Backstage alternatives fall into three real choices: build and own a framework, buy a fully managed IDP product, or choose a hybrid path that reduces maintenance but keeps Backstage at the core. The trade-off is not "free vs paid" but engineering headcount, governance maturity, time to value, and how actionable your portal is across CI/CD, IaC, and environments. The best commercial IDPs go beyond catalog and documentation.