Complete curriculum
Learn AI FinOps from first principles.
Fourteen chapters connect AI architecture, token economics, visibility, optimization and governance. Every chapter is free. Sign in with Google to start your free learning.
Start the free chaptersPart 1 · Foundations
FinOps Crash Course
Build the financial and operational foundation for managing production AI spend, using Asteria Commerce as the running case study.
Part 2 · Build AI cost visibility
From API Calls to Business Workloads
Turn raw API events into a workload catalogue and allocate shared platform costs.
AI Usage Telemetry and Cost Allocation
Normalize provider billing and request telemetry for privacy-aware showback and chargeback.
Dashboards, KPIs and Cost Anomalies
Diagnose retry storms, model changes, context growth and budget burn.
Part 3 · Optimize AI workloads
Prompt, Context and Output Waste
Measure repeated instructions, irrelevant context, oversized outputs and retries.
Caching Economics and Avoidable Uncached Spend
Calculate cache break-even points, actual savings and avoidable uncached spend.
Model Rightsizing and Workload Routing
Apply the same idea as EC2 rightsizing to AI models and routing policies.
Compare Models Without Breaking Production
Evaluate lower-cost candidates with representative data, quality thresholds and controlled pilots.
Part 4 · Advanced AI cost drivers
RAG and Embedding Economics
Optimize ingestion, chunking, storage, retrieval and model context without reducing citation accuracy.
Agent Economics and Cost Controls
Control recursive execution, tool-call amplification, context growth and multi-agent cost.
Part 5 · Operate AI FinOps
Forecasting, Budgets and AI Unit Economics
Build driver-based scenarios for growth, model mix, cache improvements and gross margin.
Governance and the AI FinOps Operating Model
Create approved-model, ownership, review, exception and savings-validation practices.