TOGETHER WITH FINOUT
AppsFlyer's FinOps Lead on Getting Every Team to Own Its AI Spend
At AppsFlyer, AI use grew from engineering to sales, legal, and procurement, and the cost had no tags and no clear owner. The FinOps team fixed it by tracking AI cost by person instead of by resource.
The results:
Every manager sees and controls their own AI spend: 40% more people started checking their costs in Finout within two months
Same day alerts: unusual AI spend reaches the right manager before the day is over, weekends included
Teams run their own AI cost reports: three budget owners built their own agents on the Finout MCP to track and report their team's AI spend, without FinOps asking
Struggling with AI costs, whether it's a tool gap or an internal ownership problem? We'd love to hear about it and show you how we approach it.
CLOUD FINOPS TOOLS
Claude FinOps Plugin Now Available
If your team keeps asking the same cloud cost questions, Cloud FinOps gives Claude a ready-made reference library.
Built by FinOps firm OptimNow and updated twice a month, it covers cloud costs, AI spend, and common sources of waste.
The library includes:
35 guides on topics such as cloud savings plans, shared Kubernetes costs, cost tags, and AI bills.
37 waste runbooks with checks, queries to run in your own account, fixes, and risks.
Search tools to help Claude find and read the right sections.
For example, a team could ask how to split shared cluster costs or check for unused disks. The key limit: this is a guide, not a tool that checks your cloud setup.
It cannot access your cloud accounts or bills, so it cannot tell you which resources are safe to delete without checks run by your team.
It needs no account credentials and stores no conversations. Price examples are dated, not live quotes.
Use it to support cost reviews, but check its advice against your own data before making changes.
CLOUD PROVIDERS
AWS cuts FSx Lustre entry cost 75%; BigQuery adds compute cost attribution

AWS
FSx for Lustre Intelligent-Tiering now costs 75% less to start. The minimum throughput and metadata IOPS are lower, so small workloads can begin cheaply.
AWS Billing adds Billing Conductor and Billing Transfer views to custom billing views. You can use them in Cost Explorer, Budgets, and the billing console.
Graviton5 R9g instances are here for Aurora and RDS. They give up to 22% better price-performance than Graviton4.
Read All AWS Updates
Google Cloud
BigQuery editions now have compute cost attribution. Billed compute costs go straight to the projects that use reservation capacity, which makes chargeback easier.
BigQuery adds project caps and principal-based reservation assignment. You can limit slots and concurrency per project so one project can't hog capacity.
Read All GCP Updates
Azure
Azure Monitor workspace auto scaling for ingestion limits is in public preview. Limits grow with your real usage, so capacity isn't fixed.
Read All Azure Updates
PODCAST
How to Optimize On-Prem Costs and Hybrid Cloud Spending
Discover how to track on-prem costs, improve resource allocation, reduce infrastructure waste, optimize GPU utilization for AI workloads, and compare cloud vs. data center spending.
Learn why unified visibility, accurate tagging, and daily financial feedback are essential for modern FinOps teams.
EVENTS
100th Editions FinOps Weekly - Hybrid Event
We’re bringing the FinOps community together for a special end-of-year event filled with inspiring talks, meaningful conversations, and new connections.
This is your chance to connect with fellow FinOps professionals, share ideas, exchange experiences, and celebrate everything our community has achieved.
Be part of this special milestone and help us make our 100th edition an unforgettable experience!
MICROSOFT AZURE
What's new in FinOps Toolkit 15
If your team spends hours tracking cloud costs, Microsoft’s FinOps toolkit 15 offers new help, but its data fixes deserve attention first.
The free, open-source toolkit adds AI support, local testing, and better tools for managing Azure discounts.
Ask AI about costs: Five agents help with cost reviews, reports, and Azure capacity. Some work without a FinOps hub; deeper trend and waste checks need one. The included hub connection is read-only.
Fix discount data: Earlier versions mixed up the fields for savings plan and reservation eligibility. Upgrade and reload past price data to correct those records. Updated reports also stop counting commitment purchases toward reported savings.
Test on your own machine: A local hub lets teams check cost data without deploying Azure resources. It needs Docker with at least 16 GB of memory available. It has no scheduled data loads or service guarantee.
Check network costs: Private hubs now support NAT Gateway for outbound access, but it adds charges.
COST ALLOCATION & TAGGING
FinOps on AWS: Why Tagging Isn’t Red Tape, but Money
Your AWS bill shows what you spent, but can you tell which team owns each cost?
Stas Kolenkin explains why tags are a basic tool for cost control, not just extra work for engineers. Tags link cloud resources to teams, products, and budgets.
That makes it easier to find unused systems, share costs fairly, and track costs per payment or user. Start with these steps:
Use five standard tags: Project, CostCenter, Environment, Owner, and ManagedBy.
Turn on cost allocation tags in AWS Billing; adding tags alone will not put them in cost reports.
Check tag coverage by spend, not resource count.
Use automatic checks to catch missing tags, and confirm tags reach new servers, disks, and other resources.
Use environment tags to help schedule dev systems off outside work hours.
Stopping servers does not stop storage charges or payments already owed under Savings Plans and Reserved Instances.
Track cost per business result, too. A rising bill may be healthy if cost per payment falls.
🎖️ MENTION OF HONOUR
AI Capacity Scarcity Shifts The Goals Of FinOps

Cutting your AI bill may not help if your team cannot get the computing power it needs.
Forrester analyst Tracy Woo argues that AI shortages are adding a new task for FinOps: deciding which teams get access, when, and for what business need.
Idle GPUs are not always wasted spend. Keeping them reserved may cost less than losing access and delaying a key product launch.
FinOps teams should focus on three areas:
Set clear rules for which projects get scarce AI resources first.
Compare each cloud provider’s reservation terms, costs, and limits on use across models and regions.
Track reserved capacity, actual use, future demand, and spend in one place.
Tracy also points to a gap in tools: few vendors show AI capacity reservations, and none can put the different units on a common basis, according to the article.
Cost control still matters. But before cutting an idle reservation, ask what losing that access could cost the business.
We Are More than a Newsletter
FinOps Learnings (30% off only for members)
Hands-on on-demand workshops led by FinOps Experts, with a limited discount for members. Learn about:
Discounts on FinOps Foundation Certifications
Use code: FINOPSWEEKLY_20 to get an instant 20% discount on the most prestigious certification bundles:
FinOps Certified Practitioner
FinOps AI Value
FinOps Technology Value
FinOps Certified Engineer
FinOps Certified FOCUS Analyst
Community
Join the Slack community where the real FinOps conversations happen
Read more from our Newsletters
FinOps Weekly Jobs Newsletter to have a weekly email with the most relevant updates every Monday
AI Economics Weekly Newsletter to know all about AI Economics and Cost Updates every Saturday
FinOps Hub
Find the right FinOps solution for your company in our Hub
Rate Today's Newsletter
Want to appear in FinOps Weekly? Talk with us







