Beyond Bots - Ilia Zintchenko, Ntropy

We talked with Nare about Ntropy and LLM's. But let's dig in more.... what is your LLM stack? How did you choose it, what were the considerations?

What are the system costs in doing this?

How do you optimize on reliability - what sort of lever are you pulling to ensure reliability?'

How are you thinking about predictive vs generative learning?

You guys have been using small and large LM's since the beginning - why is this significant?

What data sources have you been using, and are there some that are better than others?'

Have you had to scrub these data sources in any way to prep them for your LM?

What is the major benefit that Ntropy is providing by using LLM?

What would you go back and change if you could?