kfallah15

7 ms·

1. 1.▲
   by kfallah15 20d ago

   yep, it will be through enterprise licenses and our own hosted platform built on the repo
2. 2.▲
   by kfallah15 20d ago

   For the online signal, we use a LLM judge with a rubric calibrated offline by the user via TUI. UX of the calibration is a major focus area. Semantic caching is interesting, open to supporting it but not currently planned.
3. 3.▲
   by kfallah15 20d ago

   Thanks! We are going to add continual RL via Tinker soon too
4. 4.▲
   by kfallah15 20d ago

   Router and model optimization from traffic is the main differentiator
5. 5.▲
   by kfallah15 2mo ago

   Great question. You need the simulator to be realistic enough that the optimizer cannot reward hack, but it does not need to be perfectly realistic.