AI spend observability down to individual User ID's and routes

Hey HN! I'm the creator of Kalaa.

I built Kalaa because on the projects I've worked on previously, I kept having the same problem, knowing how much I was spending on AI APIs, but I couldn't really tell what was actually driving my spend. Was it a specific feature, model or User driving my spend?? Why did my spend spike on this random day ?? Also Agents you can have an agent going through a bunch of steps and making multiple calls, and when the spend goes up you're left trying to figure out which part of the process is responsible. Things got even messier when I started using multiple AI providers because I didn't have a collective place to keep track of my AI spend.

So that's why I built Kalaa a platform to help you attribute your AI API spend down to the specific model, route and user so that you know where your money is actually going daily, weekly, monthly across Open AI ,Claude and Gemini. Kalaa enables you know how much each User is actually costing you it does this by tracking individual user spend by User ID allowing you know how my much every individual user is actually costing you, paired with a per-user spend report for every user and spend alerts to enforce spend limits.

I’d love to hear how you currently track LLM spend across multiple providers and what metrics matter most in your stack. I'll be around to answer any questions!