Side Project

I’m building sntnllm.com

The idea is to build a tool that helps people who are using LLMs and text generation to keep track of their model outputs and monitor uptime. To do this, I built a few basic filters (for now, actively developing some smarter ones) and allow users to setup scheduled health checks.

I think prompt engineering is a super useful way of getting LLMs to do things, but can become a double edged sword if you don’t stay on top of it. I think this tool would be helpful if you need to rely on a model but want a good alerting system in case it outputs something unexpectedly.

I love LLMs and have worked on them at a big tech scale, so feel free to reach out with any (doesn’t have to be related to my project) questions!