This role sits within the IT FinOps team and acts as the bridge between AI technology, cost governance and business value realization. The AI Economics and Agent Engineer works closely with:
- Data Intelligence / Data DevOps within the IT FinOps team
- Platform and AI Solution Owners on architecture, model selection, deployment patterns, and cost-performance trade-offs
- Application Owners and business stakeholders on AI use cases, adoption, and value realization
- Strategic Sourcing on vendor models, commercial terms, contract levers, and optimization opportunities
AI cost visibility and token tracking
Track GenAI consumption across users, teams, applications, models, and platforms. Monitor token usage, AI credits, model calls, unit costs, and trends.
Token optimization and efficiency engineering
Reduce token consumption through better prompt, context, model, and workflow design. Promote reusable patterns, efficient agents, caching, batching, and fit-for-purpose model selection.
AI governance and guardrails
Define and maintain cost guardrails for AI usage, including thresholds, alerts, policies, chargeback rules, and consumption controls. Support responsible and financially sustainable AI adoption.
Spend predictability and forecasting
Improve predictability of AI spend by developing consumption forecasts, scenario models, cost drivers, and early-warning indicators for unexpected usage growth or inefficient patterns.
Cost-aware solution design
Influence AI solution design upfront by advising on architecture choices (run LLMs locally versus in the cloud for example), model tiers, data flows, agent patterns, and commercial implications to optimize cost-to-value from the start.
FinOps data integration and reporting
Integrate AI cost and usage data into the central FinOps platform and reporting layer. Create dashboards and insights that make AI consumption understandable and actionable for IT, sourcing, application owners, and leadership. Develop and maintain cost allocation models.
Value realization and ROI
Track ROI for AI use cases, linking consumption to measurable value such as productivity gains, avoided costs, and business outcomes. Build unit economics and foster a stakeholder community to share ideas and improve AI economics, agent engineering, and FinOps practices.
AI agent engineering
Design, prototype, deploy, and continuously improve AI agents that support FinOps activities such as anomaly detection, cost explanation, reporting automation, optimization recommendations, policy checks, and stakeholder self-service.
- AI spend under management and coverage of usage reporting across key platforms and use cases
- Improvement in cost-to-value ratio through token optimization, model right-sizing, and efficient agent design
- Forecast accuracy and reduction of unexpected AI cost spikes
- Adoption and measurable impact of AI agents used by the IT FinOps team and stakeholders
- Percentage of AI use cases with defined ROI, owner, cost baseline, and value tracking
- Good understanding of GenAI and LLM concepts, including tokens, prompts, context windows, embeddings, agents, APIs, and model selection
- Experience with AI platforms, and an ability to translate technical design choices into cost and value implications
- Strong data analytics and visualization skills, preferably with Power BI and experience integrating cost, usage, and business data
- Solid understanding of FinOps principles, cost modeling, allocation, forecasting, optimization, and performance-versus-cost trade-offs
Analytical, structured, and comfortable working with data, with the ability to turn complex consumption data into clear recommendations
- Ability to collaborate effectively with solution owners, sourcing, finance, and business teams, with a clear focus on outcomes and value realization
- 3+ years of experience in FinOps, cloud economics and IT financial management, with exposure to AI cost governance and consumption models
- High learning agility and curiosity, with a strong willingness to stay current with rapidly evolving GenAI technologies, pricing models, and optimization approaches. Learning on the job will be supported through guidance, collaboration, and hands-on experience
