Blog
Practical write-ups from my projects and internships — what worked, what broke, and the engineering decisions I'd make differently.
Why hybrid retrieval beats either retriever alone, how Reciprocal Rank Fusion quietly does the hard part, and why I now refuse to tune RAG without an evaluation harness.
Notes from my SFT/RLHF data internship at Zhipu: how training data is really made, the error taxonomy that saved my sanity, and why data quality is an engineering discipline.
How a hash grid rescued Endless Onslaught from its own particle system, why profiling beats intuition, and what game-loop optimisation taught me about retrieval indexes.
Lessons from EchoList: why tool calling beats JSON generation, how to build a graceful fallback ladder, and the unglamorous details that make an AI feature feel trustworthy.
Prefer a feed reader? Subscribe via RSS.