Writing about AI/ML, data science, and things I learn along the way.
A deep dive into Hermes Agent by Nous Research — the self-improving AI agent framework with 82k+ GitHub stars, persistent memory, and a seven-layer security model. How it compares to OpenClaw and where it falls short.
A detailed comparison of 7 AI agent frameworks — OpenClaw, NanoClaw, PicoClaw, ZeroClaw, MaxClaw, KimiClaw, and MicroClaw — covering architecture, security vulnerabilities, performance benchmarks, and which to use.
How I classified 13,000 OCR'd clinical letters into 49 categories using TF-IDF, BERT, and GPT. The biggest accuracy gains came from fixing noisy labels, not changing the model.
Why I started this blog, what I'll be writing about, and a bit about me.