Hi Hunters 👋 I'm Saad, I build solo from Italy between factory shifts. I built ProjectBlock because every AI side project I shipped had the same hole: I could see the OpenAI bill at the end of the month, but I couldn't see who burned it, or stop a single user from burning it. Adding metering meant writing the same Redis counters + audit table + Stripe glue every single time. So ProjectBlock is that layer as a service: install the Python SDK (or plug the LangChain callback), and you get per-user and per-key counters, hard caps that fire before the call goes out, audit logs, and usage-based Stripe billing. It's early and I know it. What I'd love from you: tell me where the SDK feels wrong, and what limit logic you actually need (per-team? per-endpoint? soft warnings before hard blocks?). I'm replying to everything today. — Saad

ProjectBlock

Track and cap AI spend per user, team or API key

Track and cap AI spend per user, team or API key