Presented by

Want to appear here? Talk with us

Reserved Instances and Savings Plans can cut your cloud bill by up to 45%, but only if your usage doesn't change.

Guaranteed Commitments remove that risk. For a monthly premium, Archera covers the cost of any commitment you don't use, so you can commit with confidence even when workloads shift.

Terms start as short as 30 days (not the usual one-to-three-year lock-in) across AWS, Azure, and Google Cloud. No upfront fees, no long-term contracts, just savings without the downside.

See how much you could save with a free, read-only cost analysis.

AI agents are the hot topic in FinOps circles, but real adoption is moving much slower than the conversation around it.

Most teams have not put agents to work yet, and the reasons are practical, not hype-driven. Here is what is holding things back:

The value is often unclear, since many agents just repeat what teams already know from other dashboards.

Agents need broad context, like cost, budgets, contracts and architecture, but that data is scattered across teams and locked behind security rules.

The numbers agents produce are not always trustworthy, and if something goes wrong, the FinOps team still owns the mistake.

Nobody has clearly defined what an agent is allowed to do, so giving it authority raises real questions about accountability.

An agent's own running cost is hard to predict, since it might take five steps or fifty to finish a task.

The teams let agents gather evidence and suggest actions, not make changes on their own.

They set limits through policy, not prompts, and expand trust only as results prove reliable.

Agentic FinOps works best when it earns authority step by step, not all at once.

Read More

A live conversation with me on the cloud waste your cost tooling can't see. I break down audience-submitted waste live in the closing segment.

📅 September 23rd, 2026
🕚 5:00 PM Spain / 11:00 PM ET

Kinesis now offers serverless streaming tables for Iceberg on S3, cutting delivery costs by up to 50% and query costs by up to 30%.

EC2 R9g and R9gd Graviton5 instances are now generally available, delivering up to 25% better compute performance and 30% faster database performance than the previous generation.

Lambda recursive loop detection is now live in all commercial regions, automatically stopping runaway invocation loops before they drive up your bill.

Read All AWS Updates

BigQuery Graph will require an Enterprise reservation starting April 2027, giving teams time to plan for this upcoming billing change.

BigQuery has restored daily token quotas for generative AI functions, helping teams cap unexpected spend on AI workloads.

Read All GCP Updates

Microsoft outlines four ways to lower AI agent costs in Foundry, including model routing and Batch deployments that can cut costs by up to 50% for async workloads.

Azure Resource Manager MCP now includes cost query and pricing tools by default, letting AI agents surface savings opportunities directly in your workflows.

Read All Azure Updates

Join FinOps professionals at the FinOps Weekly Summit 2026 and discover how to:

Transform FinOps from reactive cost control into strategic business value

Build scalable unit economics for the AI era

Learn proven strategies from organizations managing billions in cloud spend

📅 October 20 & 21, 2026 - 9AM ET / 3PM CEST

Learn how to optimize AWS security costs without compromising protection. AWS expert Ihor breaks down common setup errors, FinOps alignment, and budget-friendly AI security strategies.

AWS just made cost reporting a lot less messy for teams who share billing data with partners, business units, or auditors.

Cost and Usage Reports often include way more data than one team needs. Getting just the right slice used to mean building extra pipelines with Lambda, Glue, or Athena.

AWS Data Exports now lets you filter that data right at the source using simple SQL.

You pick the accounts, services, regions, or tags you care about. Only the matching rows get delivered to your S3 bucket.

No extra cleanup step needed.

Partners can now export cost data for just their customer accounts, making invoicing simpler.

Large companies can track spend for one program across many teams without extra tools.

Compliance teams can keep cost data for specific business units in its own secure location.

Service providers can send each client their own clean cost report automatically.

The setup uses a CSV file for account lists, a SQL file for the query, and one deployment script.

Updates are simple too. Change the CSV, rerun the script, and the export updates with no downtime.

Storage costs go down since only filtered data lands in S3.

Less manual work, cleaner data sharing, and lower storage costs make this a solid win for any team managing multi-account AWS spend.

Read More

Cloud bills rarely spike because of what your team is actually using. They spike because of what keeps running when nobody is watching.

A recent audit of a staging environment costing about $4,000 a month found the same problems again and again.

Idle compute: Servers and containers running all night and on weekends, even when almost nobody uses staging outside business hours. A simple on and off schedule cut compute costs by 60 to 70 percent.

These charge for both being turned on and for every byte of traffic that passes through. Free S3 and DynamoDB endpoints can remove a lot of that traffic instantly.

Without a retention policy, log storage grows forever and quietly adds up to hundreds of dollars. Setting short retention windows, then archiving older logs to cheap S3 or Glacier storage, keeps history available without the ongoing cost.

Most waste hides in idle time, unnecessary network paths, and forgotten storage settings.

Read More

A new update to OpenCost now tracks Kubernetes inference costs using vLLM and llm-d metrics.

It splits AI costs into two types: usage-based cost and allocation-based cost. Usage-based cost is what the model spends while actively working.

Allocation-based cost is what you pay just to keep the model loaded and ready, even when it sits idle.

A self-hosted model can look cheap on paper if you only count active usage.

But if the GPU sits idle much of the time, the real cost per token climbs fast.That idle time is often the biggest hidden expense.

Keeping a model warm should be a choice, not an accident.

Teams need to ask which workloads truly need instant response and which can wait or use an outside API instead.

Clear cost labels by model, team, and namespace turn a confusing bill into a debuggable one.

Your GPU bill is not a model problem.

It is a utilization problem, and fixing it means better routing, smarter scaling, and honest tradeoffs, not just picking a fancier model.

Read More

Visit FinOps Weekly Hub

Find all Jobs in our FinOps Jobs directory or join FinOps Weekly Jobs Newsletter below to have a weekly email with the most relevant updates every Monday.

Use code: FINOPSWEEKLY_20 to get an instant 20% discount on the most prestigious certification bundles:

FinOps Certified Practitioner

FinOps Certified Engineer

FinOps Certified FOCUS Analyst

FinOps for AI

Get Certified for Less

Join the Slack community where the real FinOps conversations happen