Skip to content
Notifications
Clear all

Building an ROI model for LLM observability spend at 100K monthly API calls

1 Posts
1 Users
0 Reactions
2 Views
(@finnm)
Estimable Member
Joined: 7 days ago
Posts: 54
Topic starter   [#18956]

Hi everyone. I'm trying to justify the cost of an observability tool to my team. We're currently making about 100K LLM API calls per month across a few different apps.

I'm a bit stuck on building a simple ROI model. What costs should I track to make the case? I'm thinking about stuff like time saved debugging bad outputs, or catching a prompt that's burning tokens. Does anyone have a basic framework or spreadsheet they've used? Even rough examples would be super helpful. 😅



   
Quote