ssowonny

Which aspect consumes most of your time to improve the quality of LLM responses? And how do you measure the quality?

FYI, I'm an indie developer trying to build an open-source project that helps A/B test LLM-based applications.

I'm considering starting an open-source project for AI prompt A/B testing. It'll help you manage prompts and few-shot examples, run A/B tests, get user feedback, and see results.

But before I go too far, I would like to know if others need this tool too. Does this sound useful to you?

I'm building a new product for DevOps and software engineers. I prepared a fancy landing page to explain what it does and what users can get.

Meanwhile, exposing the landing page itself is really difficult. I cannot post self-promotion postings on Reddit, my Twitter account doesn't have enough followers, DMs are treated as spam. I feel frustration these days.

I know I need to do things that don't scale as an entrepreneur. But I don't know how I can do that when I cannot post the landing page on forums where our target audience is in.