Settings

Theme

Show HN: I built Search/Recsys that learns from feedback

dna.hashgrid.ai

3 points by bicsi · 3 comments · 1 min read

Reader

Hi HN. A while ago I started building a generic people matching app based on LLM-as-a-judge. I needed a way to bypass the O(n^2) "everyone rates everyone else" trap, but generic RAG was much worse than expected.

DNA is a search-like API at the core of that product. It stores one embedding per item (like a vector DB) and gives the best match(es). You evaluate them in some way (LLM / Decision Model / users), and send the score back. This makes it improve in real time, so next time it gives you better results. In my tests, it needed 7x fewer LLM judgments than scoring down the embedding list. Paper in progress.

What I don’t know yet is what demo would make you want to try building with it. What would you want to see it do?

1 thread
teom

Looks really cool! Did you also publish a whitepaper on how it works behind the scene?

  • bicsiOP

    We are going to publish our experimental research soon. Meanwhile feel free to contact me directly is you want to be up to date with all the info, feel free to join the Discord community: https://discord.gg/jFQKUvZ4k

Keyboard Shortcuts

j
Next item
k
Previous item
o / Enter
Open selected item
?
Show this help
Esc
Close modal / clear selection