Finout

San Diego, we're coming for you. 🌴 We're throwing the biggest party FinOps X has ever seen, turning up the heat with the swag and well... you'll just have to stay tuned for all the surprises on the way 😉 Make sure to register to party with us here: https://hubs.la/Q04gC9Zx0 #FinOpsX

Look at what has shipped or been announced since February: → ServiceNow unveiled Action Fabric at Knowledge 2026 (May 5–7). Per-agent-action billing. → Workday CEO Aneel Bhusri publicly said agent metering offers "a lot of upside." Their version is coming. → HubSpot is reportedly building the same access-metering layer. → SAP went the other direction and blocked unauthorized agents from its APIs in April. Same underlying force, different mechanic. → Datadog capped its MCP server at 5,000 daily, 50,000 monthly agent requests. → Cursor restructured to four billing axes in a single May 2026 release. → GitHub Copilot moves to AI-credit overage billing June 1, 2026. → Anthropic locked third-party tools out of subscription tiers on April 4, 2026. → AWS and OpenAI consolidated agent traffic into Bedrock in late April. That is not eight vendors moving independently. That is your entire AI bill. The reason is structural. The seat-based pricing model that defined enterprise software for two decades cannot pay for what AI agents do. Are you ready for what's coming?

Ever since Anthropic moved to usage-based pricing, FinOps teams have been asking the same question: "Who's actually driving our Anthropic spend?" Until now, that data wasn't available for Claude.ai. It is now — and Finout supports it. Per-user Claude.ai spend now flows directly into MegaBill, alongside your AWS, GCP, Azure, and Kubernetes costs. What you get: → Named-user attribution across Claude chat, Code, and Cowork → Breakdowns by model and token type → Allocate, showback, and detect anomalies — just like you do with your cloud spend Read more: https://hubs.la/Q04gfwSr0

The FinOps tools built for the pre-AI era are falling behind, and the infrastructure is simply moving too fast. Finout Extend 2026 puts the seven companies solving these problems on one stage, and highlights how Finout stars at the center. June 4, 2026 12:00 PM to 2:15 PM EST \ 9:00 AM to 11:15 AM PST \ 5:00 PM to 7:15 PM BST Register: https://lnkd.in/gVPEQPxW Feat: CircleCI, Sawmills, Seemore Data, InfrOS, Portkey, Alvin, TwoDelta #FinOps #Extend2026

If you manage Claude at your company, you've probably had to deal with a frustrating question: who is actually driving the bill? Until last week, that was hard to answer. Anthropic's old Admin API could tell you a workspace spent $40K on Opus last month, but it couldn't tell you which people, on which teams, doing what work. Their new Enterprise Analytics API changes that. You can now pull per-user cost data across Claude, Claude Code, Cowork and the Office agents, with breakdowns by model, context window and inference region. So instead of "our Claude bill went up 30%," you can say "12 engineers on the platform team drove most of the increase, mostly through Claude Code sessions in the 200K-1M context window." That's a conversation you can actually have with a team lead. A few things stood out to us as we dug into it. The context window breakdown is more useful than it looks. When Anthropic removed the 2x pricing premium on long context back in March, the natural guardrail on context size went away with it. Per-user visibility is now the main way to keep an eye on whether people are running expensive long-context sessions when they don't need to. Claude Code is finally trackable at the human level. It used to be a black box because it runs locally and only showed up as API key activity. Now you can actually look at cost per commit, cost per PR, cost per session, and compare productivity to spend. The bigger pattern is that AI providers are moving through the same cost visibility maturity curve cloud went through, just compressed. What took AWS eight years is happening with Anthropic in roughly twelve months. We wrote up a full breakdown including what's still missing from the API and a four week playbook for getting started. LINK HERE: https://hubs.la/Q04fZWh80 #FinOps #AI #Anthropic

🎟️ 🎡 🍭 🍿 🎪 🎯 🎠 📸 🎶 🎟️ 🎡 🍭 🍿 🎪 🎯 🎠 📸 🎶 🎟️ 🎡 🍭 🍿 🎪 🎪 Welcome to The Finout Carnival 🎪 FinOps after Dark is back at FinOpsX and we're going bigger than we ever have. We're taking over the Coronado Terrace at the Marriott Marquis on June 9th from 7-10 PM and turning it the real reason you flew to San Diego. Lights, carnival games, top-shelf cocktails, and basically every person you've been meaning to DM on LinkedIn for the past six months, all in one place. If you're at FinOpsX, this is your move that night. Reserve your spot before the cap hits 👇 https://hubs.la/Q04fQZTC0

80% of companies are missing their AI cost forecasts by more than 25%. The fallout? 84% say it's eroding their gross margins. 😬 Translation: most teams have no idea what AI is actually going to cost them next quarter. Asaf Liveanu breaks down the biggest AI cost drivers to watch in 2026. Read it before your next finance review: https://lnkd.in/eUAm2Y5t

Your AI assistant — Claude, Cursor, and more — can now connect directly to your Finout cost data and answer any FinOps question in plain language. Not raw billing data. Your actual allocation model — virtual tags, shared cost rules, cost centers, and all. Ask things like: "What did Platform Engineering actually spend across AWS, Snowflake, and Databricks this month, after shared cost reallocation?" "What drove last week's cost spike, and who owns it?" "What's the unit cost of serving our top 10 customers?" "How are we tracking against the financial plan, and what's our commitment utilization?" No dashboards. No exports. No waiting on the FinOps team. Just accurate, allocated, real-time answers — secured by the same role-based access controls as your Finout platform. This is what agentic FinOps looks like. https://lnkd.in/eJishAwb #FinOps #AgenticFinOps #MCP #CloudCostManagement #AI #CloudCost #AIAgents #FinOpsX

The hardest problem in FinOps isn't visibility. It's indecision. Engineers see re-architecture risk. Leadership sees an unclaimed number. Both sides have valid fears, and whose fear wins usually decides whether anything actually changes. On May 6, Marit Hughes (Deloitte) and Carlos Galán (Finout) are getting into it: → Why "cutting 5% while unit cost rises" isn't a FinOps win → How to translate between engineering, finance, and product → How to report value for the bad decisions you didn't make Save your seat: https://hubs.la/Q04fkgc60

Waste reduction isn't always what the business needs from FinOps. The bigger work: unit economics, decision quality, and defending the savings you chose not to pursue (or claiming wins for the bad decisions your team prevented). Marit Hughes (Deloitte) and Carlos Galán (Finout) get into it on May 6. https://hubs.la/Q04dTQ2n0