> ## Documentation Index
> Fetch the complete documentation index at: https://narev.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# FinOps for AI Framework
> A practical three-step FinOps framework to measure, track, and optimize LLM spending in production without compromising quality or user experience.
**LLM optimization isn't about spending less—it's about spending smarter**. While some teams burn through their AI
budget delivering basic features, others build sophisticated experiences for a fraction of the cost. That efficiency
gap compounds fast. Don't be on the wrong side of it.
You landed here because someone raised concerns about the cost of your chatbot, agent, or workflow, and now you need a plan.
Common reasons teams start exploring FinOps for AI:
* **Unit economics don't work** - you're paying more to serve customers than they're paying you
* **Unexpected cost spikes** - everything was fine until your provider's invoices started flooding your inbox
* **Proactive optimization** - you want to build efficiently from day one (kudos for that!)
No matter what brought you here, there's good news. LLMs can deliver exceptional results **and** cost less than the current spend.
## Start with the foundation
You might want to jump straight to optimization, but here's the truth: **you can't optimize what you can't measure**. Use a structured approach that has been battle-tested across dozens of organizations:
1. **Know your objective** and who's in charge
2. **Know what you're spending on**, down to a feature and responsible team
3. **Optimize your apps**, one use case at a time
Walk through each step:
* [Step 1: Know your objective and who's in charge](/guides/finops-for-ai/step-1)
* [Step 2: Know what you're spending on](/guides/finops-for-ai/step-2)
* [Step 3: Optimize](/guides/finops-for-ai/step-3)