Show HN: AI Getting Started – A template helps you start a AI SaaS with ease

github.com

20 points·timqian··6 comments

Given that there's plenty of options for every point in the README.md, one thing missing is how to guarantee that your stack does not miss requests from paying customers, metering usage & avoid ballooning server costs. I see a lot of YC startups trying to solve this Lago, Paigo etc.

I'm trying to evaluate best serverless solutions for inference without compromising on client usage & reducing idle time on GPU boxes. So far its down to base10, HF, Banana, I'll end up pooling them all & then sending requests between them. For dedicated training boxes Lambda, Modal, Oblivus, Runpod are the contenders.

If it's a small startup I don't think you can beat Replicate as far as costs.

To track usage you need a credits system. Basically you have to read a number from a file, subtract the cost for the operation, and if it's less than zero throw an exception. It does take a little bit of development, but I don't think you need a whole other startup to handle that. You can do the core part of it for one type of operation/cost in less than a day. Maybe a few days to debug something.

Is there a reason not to use Supabase Auth if you use Supabase anyway?

We did not use supabase pg-vector at first (using pinecone)

That makes sense. Thanks.

are they trying to sell products of their porforlio companies?

Reply on news.ycombinator.com