Posts
All the articles I've posted.
-
Visual UIs Are Now Possible in MCP Servers
MCP servers can now render interactive UIs directly in Claude Desktop's chat window. Not just text responses—actual HTML with JavaScript, maps, charts, anything.
-
Did You Know CLIP Works as an AI Image Detector?
OpenAI's CLIP model was trained to match images with text descriptions. But it also works remarkably well at detecting AI-generated images. No fine-tuning required—just extract embeddings and add a simple classifier.
-
From Prototype to Production: Building a Multimodal Video Search Engine
Expanding an afternoon hack into a production-like multimodal video search engine with CLIP, Whisper, and ArcFace — containerized with Docker and powered by Postgres+pgvector.
-
I Stacked 3 Small ML Models and Got Video Search That Feels Like Magic
Combining CLIP, Whisper, and ArcFace into a stacked video search prototype. Each filter alone returns roughly right results, but stack two or three and the precision jumps dramatically.