Hey guys how are you doing today? Trying to get a realistic picture across different products. I keep seeing two extremes: people spending ~$50/mo and people quietly bleeding $5-10k+/mo on inference. So I'm collecting data points. If you've got AI/LLM features in production: roughly what's your monthly AI bill? what's driving most of it (which feature or model)? have you done anything to bring it down, and did it actually work? I'll share what I'm seeing in the comments. Lets...