FinOps for AI · Level 2

You can see your AI spend. Now learn to shape it.

Level 1 answered “what did we spend, and who spent it?”. Level 2 answers the CFO’s question: “what will this cost us next year, and should we?”. Policy and governance, planning and estimating, forecasting, budgeting, and how the FinOps practice itself adapts to AI, built on the FinOps Framework capabilities for AI.

The Level 2 roadmap
01
AI Policy & Governance
Guardrails, owners, the AI investment council
02
AI Planning & Estimating
A first number for every AI idea, before the money is spent
03
AI Forecasting
From straight lines to driver-based, collaborative forecasts
04
AI Budgeting
Rolling budgets, re-approval in-cycle, value not just variance
05
Adapting the FinOps Practice
Tools, assessment and operations for a fragmented AI bill

Seeing AI spend is not the same as controlling it.

The $400 pilot that costs $12,400 in production
Proofs of concept skip retraining, storage, logging and licences. Nobody estimated what it would cost at scale.
Forecasts drawn from last quarter
A straight line from $75K says $82K. New teams, new habits and a retrain make it $118K.
Budgets that stop meaning anything by February
A flat annual line cannot follow a launch, a retrain and a price cut. Teams stop looking at it.
What you'll learn

From visibility to strategy: the five capabilities that shape AI spend.

01
AI policy & governance
Policy owners and doers, trust-but-verify versus control-first, guardrail policies, the AI investment council, governance KPIs.
02
AI planning & estimating
Why AI is hard to estimate, the six-step estimate, POC-to-production, estimate variance and the practices that make it a discipline.
03
AI forecasting
Granularity, collaboration and forward-looking inputs, the six-step lifecycle, scenarios, and forecast accuracy KPIs.
04
AI budgeting
Adaptive budgets, in-cycle re-approval, budgets tied to business outcomes, rolling practices and budgeting KPIs.
05
Adapting the FinOps practice
Tooling for a fragmented AI bill, FOCUS, build or buy, assessing and operating the practice, practice KPIs.
97 short lessons. A quiz after every chapter. One timed review quiz.
Self-paced · Progress saved on any device
Serge Niagne
Taught by Serge NiagneCertified in FinOps for AI, FinOps Practitioner and FinOps Engineer. Solution architect for 15 years, Microsoft Certified Trainer, 11,000+ students taught.

Questions

Does the course include the official FinOps for AI certification?
No. The official FinOps Certified: AI Value credential is awarded by the FinOps Foundation through its own training path and exam on learn.finops.org. This is an independent Level 2 course built on the FinOps Framework capabilities for AI, focused on the practices you apply at work.
How is the course organised?
An introduction and five chapters, 97 short animated video lessons in total, with one running scenario (Calloway Retail) and KPIs in every chapter, a 10-question quiz after each chapter and a timed 50-question review quiz at the end. It is self-paced, and your progress is saved on any device.
What do I need to know first?
Level 1 or equivalent experience: you should be comfortable with allocating, ingesting and reporting AI spend. You do not need a machine learning or data science background.
Do I get a certificate?
Yes. When you mark every lesson as complete, you can download a Cloud Institute Canada certificate of completion with your name and the course title, ready to add to LinkedIn. It is separate from the official FinOps Certified: AI Value credential, which the FinOps Foundation awards.
How long do I have access?
For life, including future updates to the course.
Can I get a refund?
Because the course is digital and access is immediate, purchases are final once access is given, except where Québec consumer protection law provides otherwise. If anything is wrong (you cannot access the course, it is not what was described, or you were charged in error), email support@cloudinstitute.ca and we will put it right. Details in our Terms.
Can I enrol several people from my team?
Yes. For group access, see For Teams or write to support@cloudinstitute.ca.