Operations | Monitoring | ITSM | DevOps | Cloud

AI cost monitoring is the continuous tracking of AI and LLM spend in real time, broken down by the models, features, teams, and customers generating it. It is not the same as reading the monthly bill - done well, it shows spend as it happens, flags anomalies before they become invoices, and connects every dollar to an outcome so finance can protect AI ROI instead of explaining it after the fact.

Agentic AI cost is what you pay to run AI agents, and it is mostly tokens. An agent does not answer once. It loops, calls tools, reads the results, and reasons again, re-sending a growing context every step. Anthropic found agents use about 4x the tokens of a chat, and multi-agent systems about 15x. You control it by capping runs, right-sizing the architecture, routing, caching, and measuring cost per task, then tying every agent to the AI ROI it produces.

Enterprise networks have outgrown the flat, predictable topologies they were built on. Traffic now crosses on-premises data centers, multiple public clouds, SD-WAN links, and the home connections of a distributed workforce. When any one of those segments degrades, the business feels it as slow applications, failed transactions, and frustrated users — long before anyone confirms the network is the cause.

In May I wrote that you probably don’t need private PKI for internal infrastructure, because DNS validation gets a publicly trusted certificate onto hosts that never touch the internet. In June I pointed out that Apple enforces an 825-day cap on private certificates, so your own CA doesn’t even free you from the browser vendors. Two days ago I told you that public client certificates stop renewing in October, and that the replacement is a private certificate authority.

Observing fast-growing agentic workloads is no small feat, especially if you try to build your own monitoring stack or rely solely on tools built for a time before LLMs. At Grafana Labs, we know this all too well.

If you haven’t heard about our MCP server, you should really check it out. It’s probably the best way to give your agents access to the power of GitKraken’s integrations and features. Our MCP tools also help your agents understand your codebase in a way that we think lowers your token usage and improves their output.

Most PR descriptions leave out the one thing a reviewer actually needs: why the change was made. And AI review bots that live natively inside GitHub tend to solve that with noise instead, dropping comments a reviewer then has to sort through to find the two that matter.

Learn how Megaport Webhooks deliver real-time events to your tools, enabling faster responses, automated workflows, and less manual work. The tools that run modern infrastructure operations, SIEM platforms, IAM solutions, and CI/CD pipelines are built to act on events as they happen. But the problem with this approach is that finding out what has changed on your account has always meant asking by either polling the API on a schedule, or waiting for an email response.

Summer is the perfect time to step back from the alert stream and revisit the ideas your whole monitoring practice quietly rests on. Not the newest tool or the shiniest dashboard — the fundamentals: what to measure, how to read it, when to wake someone up, and what to do when something breaks at 2 a.m. We packed six of those fundamentals into a free, no-signup PDF — The DevOps & Monitoring Summer Workbook, complete with hands-on exercises and an answer key.

In April, Anthropic disclosed that its new AI model, Claude Mythos, had found a vulnerability in OpenBSD that survived 27 years of human review. Mozilla used the same model to find 271 security flaws in Firefox in a single evaluation pass.