Operations | Monitoring | ITSM | DevOps | Cloud

Eight new Azure recommendation types now scan your environment for idle, unattached, and over-provisioned resources, then tell you exactly what to cut.

This is the first in a series we’re calling AI ROI Dispatches, where we share stories from CloudZero and our customers on tying AI spend to real business outcomes.

CloudZero now splits an Amazon Bedrock line item into spend by model and by the team, user, or project behind each call, straight from your AWS bill.

How to track expenses for a business: categorize expense types (operating, software, cloud, travel, capital), choose a tracking method (spreadsheet, accounting software, expense management tool, or cost intelligence platform), connect data sources (bank feeds, cloud billing APIs, SaaS invoices), assign ownership per cost center, set a reporting schedule, and audit quarterly.

The CloudZero VS Code Extension 2.0 adds a sidebar that brings CostFormation namespace management directly into your editor, so you never have to break flow for a browser tab.

In Explorer, you build a filter set and group-by to answer a cost question, and often that’s exactly the configuration you’d want to save for later. But saving it as a View meant navigating away from Explorer, opening the Views page, and rebuilding the same configuration from scratch: filter by filter, dimension by dimension. That friction was enough to discourage saving exploratory analysis as a View at all You can now save any Explorer analysis as a View in place.

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.

Cost monitoring, exploration, investigation, and reporting used to live in separate places, so following a single cost question meant hopping between tools.

Your AI calls already emit OpenTelemetry: your LLM gateway exports it, and it’s the open standard your own services can speak. But you don’t have anywhere to turn those spans into spend you can allocate to an outcome. Now you can. CloudZero exposes an OpenTelemetry endpoint that doesn’t care what’s on the other end.

You can already split AI spend by team and by model. But that’s not what your CEO asks in the QBR. The question is what you got for it: what did it cost to ship that feature, to launch that campaign, to serve that customer. And is the AI bet behind it paying off? Now you can allocate AI spend to the outcomes you own: customer, product, feature, the strategic bet on the P&L. Not just the team that spent it.