The illustrations were created in cooperation between humans and artificial intelligence. They show a future in which technology is omnipresent, but people remain at the center.
AI-generated illustration
FinOps and Token Management: Cost Control for Cloud and AI Services
Manage budgets, usage-based costs, and responsibilities company-wide
Online
2 days
German
€ 1.290,-
plus VAT.
€ 1.535,10
incl. VAT.
Booking number
43282
Venue
Online
.svg)
5 dates
€ 1.290,-
plus VAT.
€ 1.535,10
incl. VAT.
Booking number
43282
Venue
Online
5 dates
plus VAT.
incl. VAT.
Booking number
Venue
0 Dates
Become a certified
Machine Learning Engineer
Machine Learning Engineer
This course is part of the certified Master Class "Machine Learning Engineer". If you book the entire Master Class, you save over 15 percent compared to booking this individual module.
To the Master Class
In-house training
In-house training just for your employees - exclusive and effective.
Inquiries
In cooperation with
In cooperation with
In the past, IT and procurement made decisions based on costs; today, usage determines costs. Cloud services are billed based on consumption, as are AI assistants and agents: Every request, every document analysis, and every agent run costs money—the moment someone clicks “Send.” This shifts cost responsibility across the board: A product owner determines ongoing costs through the architecture of their application, while a clerk incurs costs by running an agent without even realizing it. Traditional annual budget planning isn’t designed to handle this.
For the cloud, there has long been a proven solution: FinOps. This two-day onlinetraining applies this approach to AI and agents, making it applicable to everyone who plans, is responsible for, and actually incurs costs on a daily basis.
You’ll understand the cost mechanics behind tokens, credits, and agent runs; calculate costs and ROI for a specific use case; set up budgets and allocations at the group and employee levels; and learn how to foster cost awareness among your workforce without hindering the use of AI.
Contents
1. The Shift in Cost Responsibility
- From Order to Consumption: Why Costs Are Incurred in Operations Today.
- Why usage-based billing undermines traditional budgeting and approval processes.
- Who actually incurs costs today: application managers, departments, individual employees.
2. Why FinOps Was the Answer to This Trend
- The FinOps Foundation's FinOps Framework: Principles, Roles, and the Interplay Between Finance, Business Units, and Technology.
- Inform, Optimize, Operate: Cost Data and Allocation, Effective Leverage, Budget Planning, and Forecasting.
- FinOps as an ongoing process rather than a one-time project.
3. AI Turns Every Job into a Cost Center
- Where AI costs arise: applications, workplace assistants, and background automation.
- Two user groups, two cost models: direct API access vs. assistants and agents in the business unit.
- Licensing plus usage as the new normal: the blind spot of unbudgeted AI use.
4. Agents in the Company: The New Colleague
- What distinguishes an agent from an assistant: the ability to perform tasks independently, access to tools, and the ability to run unsupervised.
- Clarify ownership: Who is responsible for an agent's technical aspects, who manages their budget, and who deactivates them?
- The cost difference between a well-built and a poorly built agent, illustrated by an example.
5. Understanding Tokens: The Cost Mechanics of AI Usage
- Token as a billing unit: Input, Output, Context Window, Reasoning Shares, Credits.
- Everyday Cost Drivers: Model Selection, Prompt Length, Document Context, Agent Loops.
- Effective strategies: model routing, prompt caching, batch processing, context limiting.
- Calculation Exercise: Calculate costs per process, per employee, per department, and per month.
6. Token and Agent Management at the Group and Employee Levels
- Structured Approach Instead of Flying Blind: API Key Governance, Gateway Level, and User Groups as a Framework.
- Take inventory of agents, assign them to the appropriate teams, and set budgets and limits for each group and individual.
- Allocation to departments and forecasting: extrapolating from the pilot group to the rollout.
- Exercise: Calculating ROI for an AI use case—saved labor time versus operating costs.
7. Raising Awareness at colleagues: Drawing Attention and Shaping Communication
- Why Bans and Restrictions Fail—and What Works Instead.
- The key message: AI is billed per use, not per year.
- Transparency as the most effective tool: Providing usage and cost feedback to teams rather than issuing individual warnings.
- Shaping Communication Within the Company: Occasions, Channels, Tone, and Influencers.
8. Overall Governance: Roadmap, Roles, and Business Case
- Roles and division of labor among IT, Purchasing, Controlling, business units, and Compliance.
- Select an operating model: centralized unit, federated model, Center of Excellence.
- Develop a 90-day roadmap and business case and present them to management.
- Templates for cost control, as well as recommendations for appropriate tools, dashboards, and reporting processes.
Your benefit
- You understand why costs are incurred at the time of use rather than at the time of ordering, and you can contextualize and explain this trend within your organization.
- You understand the cost mechanics behind tokens, credits, and agent runs, and you can calculate AI costs per task, per person, and per department, and budget for them proactively.
- You calculate the ROI for a specific use case and can explain which AI applications are cost-effective and which are not.
- You know how to treat agents just like employees: keep track of them, assign them responsibilities, provide them with a budget, and ensure clear ownership.
- You set up allocations, budgets, and limits at the group and employee levels and use AI to make costs transparent based on their source.
- You know how to foster cost awareness among employees and structure communication around this topic without hindering the use of AI.
trainer
Philipp Garra
Methods
The course consistently focuses on concrete arithmetic examples rather than abstract models:
- Calculation Exercise for Estimating AI Costs per Process, Person, Department, and Month
- ROI calculation for a real-world use case, comparing time saved to ongoing token and operating costs
- Templates for Cost Control, Budgeting, and Business Cases
- Case Studies on Well-Designed vs. Poorly Designed Agents and Their Impact on Costs
Prerequisites: Prior knowledge of cloud computing or controlling is helpful but not required.
Final examination
Recommended for
- Budget managers: ITcontrollers , and cost center managers who are responsible for preparing and defending budgets
- Software Engineers: developers and Tech Leads who want to monitor the cost side of cloud and AI services
- Product owners and business analysts, whose decisions regarding feature sets, model selection, and the degree of automation determine operating costs
- Business users of AI who incur costs for every transaction in their day-to-day operations
- IT Procurement and IT Purchasing
- IT License Managers
- IT Service Managers and Service Owners
.svg)