This is our 3rd pivot and I just realized I never shared it here.

We built Latitude for teams who collaborate on prompts and want to iterate with confidence.

We've seen people use spreadsheets, notion docs, and even build internal tools to manage prompts, but it's still pretty much a manual workflow based on vibes.

With Latitude you get:

1. A prompt manager to enable non-technical folks to create, edit, and test prompts 2. A suite of evaluations (llm-as-judge, human in the loop, etc) to understand how different prompts perform

All of it powered by real production data captured as logs that you can turn into datasets for testing.

We already have a few hundred users and the feedback has been great, we're a small team and we ship fast (proof: https://docs.latitude.so/changelog), so we're looking for more feedback and ideas.