If we are getting general purpose classifier, decision engine, probabilistic inference machine and likely a couple of other use cases for the fraction of latenc

If we are getting general purpose classifier, decision engine, probabilistic inference machine and likely a couple of other use cases for the fraction of latency/time/price of using SOTA LLMs for such tasks or training specialized classifiers this is very interesting project. Considering in future we would be able to supply it with our own facts knowledge bases and fine-tune for specific domains, this enables a lot of interesting use cases in various software domains.