Datadog’s FinOps analyst says AI cost management starts with tagging and model governance

What changed

Datadog’s senior FinOps analyst Deeja Cruz puts a clear spotlight on the realities of managing AI costs in enterprise environments. While AI introduces new technical layers and a fresh taxonomy, the foundational discipline in FinOps remains unchanged: know what you’re using, why you’re using it, and what it costs. Cruz stresses that proper tagging and model governance are the starting points for any AI cost management effort. Without them, tracking expenses across AI models and cloud resources becomes a guessing game.

Why builders should care

AI workloads shift cost structures and accountability within organizations. Unlike traditional cloud resources, AI requires tagging not just by application or environment but down to the specific models in play. Governance around model usage and deployment ties directly to controlling runaway costs and avoiding budget surprises. Builders and operators who don’t bake in tagging and governance from day one will struggle to isolate cost drivers and justify AI investments. This could slow AI adoption or force budget cuts.

The practical takeaway

Tagging AI models and maintaining rigorous model governance create transparency in a complex cost landscape. This approach forces clear ownership and usage tracking, which simplifies budgeting and cost optimization. Enterprises can leverage these FinOps basics to hold teams accountable, identify idle or inefficient models, and optimize spend across AI and cloud resources. Ignoring this increased granularity risks inflated AI bills and fractured cost accountability.

What to watch next

Expect AI FinOps tooling and best practices to evolve rapidly around tagging standards and model governance frameworks. Cloud and AI platform vendors might start enforcing or simplifying model-level cost tagging. Enterprises will likely demand solutions that integrate model metrics with existing FinOps workflows. Watching how this discipline matures will reveal which companies manage to keep AI experimentation affordable and scalable versus those that face uncontrolled expense growth.

AI Quick Briefs Editorial Desk