Vineetyadav2

well I am building a verification layer for hospitals and security systems

we added mechanical layers where ever possible , Think of it as you wanted a piece of quality work who will you choose A single person or 4-5 diff people who cross check each other work

Thats the architecture we follow here but its not just we just took multiple model because in order to verify answers we had to implement some layers, Like for a code query it goes to diff layer, for reasoning , for summary, for factual and so on We tried to build this as mechanical as we can because a single Llm can't see its own mistakes unless someone points that out .

I feel like AI companies are just implementing confident answers not a overall good answers

I was not aware about this existing thing but our purpose is whole diff

I kept getting confidently wrong answers from AI models, even when they were wrong . So I changed the approach Imagine you want to get a quality piece of work whom will you choose a single person or multiple people who cross check each other work I build this as multi model orchestration so every answer gets check before you see it

yes Experience is something that is better than continuous training on same data A AI need training over 10000 times just be average at playing chess while humans need few hundred time because do teach things that continuous training can't