Wes Kennedy
Systems archaeologist by instinct, technical marketer by trade.
Series
LLM Inference - How It Works
What actually happens when you send a message to an LLM — from tokens to attention to output, one rabbit hole at a time.
32 articlesLLM Training - How It Works
What actually happens when you train a model from scratch — from data to architecture to a 16,384-GPU training loop, one rabbit hole at a time.
47 articles