Engineering teams are shipping faster than ever. AI coding tools like Claude Code and OpenAI’s Codex have quietly removed some of the biggest friction points in the development cycle — and the result is that FinOps teams are being asked to keep up with a pace most practitioners haven’t fully reckoned with yet.
That acceleration has a cost consequence. More shipping means more services, more experiments, more infrastructure spun up without review cycles. The volume of “what is this and why does it cost that much?” questions landing on FinOps teams is growing, and the old workflows of manually pulling the data > building the context > writing the response just aren’t keeping up.
Since we launched our Claude Code plugin last week, a theme is emerging from conversations with customers: the practitioners who are adapting fastest aren’t just using AI to answer questions more quickly. They’re rethinking what their job actually looks like going forward.
When the mechanical work gets handled automatically, what’s left are the strategic parts of the role — influencing decisions, building accountability across teams, connecting cloud spend to business outcomes through unit economics. The parts that are actually fun. Not to sound like a cliché, but that’s working smarter, not harder.
What ‘filling the gaps’ actually looks like in practice
The most common feedback from people who’ve started connecting their cost management systems into AI platforms like Claude is some version of: “I continue to be amazed at how it’s filling in gaps that were previously hard to overcome.”
If you’re currently in a FinOps role, stop and think about your typical week. How many one-off questions land in your inbox or Slack?
- “What was our service cost last week?”
- “Why did we have this spike?”
- “How do I build a report that shows X?”
Say each one costs you 20–40 minutes. Multiply that across a year and you’ve spent a meaningful chunk of it as a human query interface.
When you connect cost intelligence into Claude via MCP, those questions get answered directly, often before they reach you. Cloud cost management shifts from a reactive function to a self-service layer embedded in engineering workflows. Engineers can ask about cost anomalies in the tools they’re already working in, get context-aware answers grounded in actual usage data, and self-serve in ways they simply couldn’t before. Cost visibility shifts left, closer to where the spending decisions are actually made.
The GitHub example
Here’s where it gets interesting for the people willing to experiment.
Imagine you’re trying to understand why a particular service’s costs jumped 30% this sprint. In the old world, you’d pull cost and usage data, then separately dig through GitHub to see what shipped, then manually try to correlate the two, and maybe set up a quick call with an engineering manager. A good FinOps analyst might crack it in a few hours. A great one who knows exactly where to look might do it in 45 minutes.
Now imagine Claude with CloudZero and GitHub both connected as MCP servers. You ask: “What shipped this sprint and how does it correlate with our cost increase?” Claude pulls recent commits and PRs from GitHub, cross-references them with actual cloud cost optimization trends in CloudZero, and builds you a coherent picture in seconds. Not an approximation. Actual data from both systems, synthesized.
That’s not replacing a FinOps analyst. That’s a FinOps analyst running at 10x speed. The analyst still has to know what question to ask, how to interpret the answer, what follow-up matters, and how to translate it into a recommendation leadership will act on. The judgment layer remains entirely human. The drudgery layer gets automated.
The practitioners who figure this out early will become the people who can close investigations in minutes instead of hours. They’ll be the ones who show up to the quarterly business review with insights nobody else saw coming. They’ll look, as one customer recently put it, “like a hero.”
You will look like a hero. Or you will look like the bottleneck.
This is the part most people in our space don’t want to sit with.
As we’ve been talking to practitioners who are seeing these AI-assisted workflows for the first time, one question continually comes up: “What happens to my job if it can do all that?”
It’s the right question, and the honest answer is nuanced.
If your value to the organization is primarily answering one-off questions, running monthly reports, and reviewing dashboards, that’s probably at risk. Not because AI will eliminate the role, but because AI will eliminate the need for a human to sit between the data and the person asking about it.
If your value is strategic, like understanding tradeoffs, building relationships with engineering leaders, influencing architectural decisions, and driving accountability across teams, AI makes you better at that job by handling the mechanical work you currently do to get there.
The people I genuinely worry about in any role at a SaaS company right now aren’t the ones asking “what happens to my job?” It’s the ones not asking the question at all.
The people moving fastest right now aren’t waiting for a formal initiative. If your current cost management platform supports MCP integration, start there. Connect it to Claude, spend a week asking it questions you’d normally answer yourself, and see what happens.
If your cost management platform doesn’t support that kind of integration yet, that’s worth knowing. The ability to connect your cost data into AI cost management workflows isn’t a nice-to-have anymore. It’s becoming a baseline expectation for what a modern FinOps platform should do.
FinOps is in year 13. AI is on day one.
We’re at the very beginning of figuring out what AI-augmented FinOps practice actually looks like. The patterns aren’t set. The best use cases are still being discovered by practitioners willing to tinker.
We’re already hearing from people who were skeptical — worried that automating the routine work meant automating themselves out of relevance. What they’re finding is the opposite. The hours they used to spend on ad hoc requests and report-building are going toward the work that actually builds a career: strategic conversations, proactive analysis, a seat at the table they couldn’t get to before because they were too busy answering questions.
That’s the trajectory we’re seeing, and it’s available right now to anyone willing to take the first step.
The Cloud Cost Playbook
The step-by-step guide to cost maturity
