alright wow building a RAG component is actually a mindblowing experience.

👨‍💻 /uno and /outside, @unofficial

Building something new. 🟪 Host: /venturecapital /adventures /theitinerant

building out the discovery engine @unofficial promised during the demo day!

a search of “NYC marathon runners” should display these very quickly:

1. social content
2. communities
3. media content

Which GAN foundational model are you using?

not using a GAN! — we use Gemini Pro as the LLM in the flow of data below.

for semantic search we use a different model (i didn’t build that particular embedding flow)