dex

@dexhorthy

San Francisco, CA

Pinned
dex
@dexhorthy
Jun 17
At HumanLayer, we’re on a mission to solve the AI slop code problem. In 2025 we open-sourced our Research, Plan, Implement framework, now deployed inside fortune 500s like Block and Uber - places where shipping slop is just not an option And that was just the beginning. Today,
00:00
219
dex
@dexhorthy
2h
if you're gonna be at AGNTCon in Amsterdam next week, lets catch up - I'll be talking software factories, slop code benchmarks, and why the future of Agentic Dev will be built from composable building blocks hmu for a discount code if you want one, and if you're building
00:00
4
dex
@dexhorthy
16h
why are models bad at harness engineering!?
dex
@dexhorthy
16h
Replying to @martin_casado and @cansar
I think harness / pipeline / prompt engineering are some of the hardest things to get models to do well. Part because small training set, part because slower more expensive RL oracle if you’re building on inference, and part because a lot of the recent training data is ai
50
dex
@dexhorthy
19h
Many are saying
14
dex
@dexhorthy
Sep 5
Milestone - /show-me has over 11k installs and 2.5k github stars!
15