Writing(3)
On AI and LLMs, engineering, and how software actually gets built.
Generic vs Domain Specific Tokenizer Eval
Why a small tokenizer trained on your own domain can cover the same text in fewer tokens than a general purpose one many times its size.
LLM Does Not Know Your Business
When ChatGPT was launched in 2022, it was more than just another product. It was the beginning of a new chapter in technology. Some people immediately embraced it, while others…
What Separates a Good Data Product From a Great One
Stability, velocity, and maturity usually describe systems and architecture. Here, I am applying them to data. For any company that builds a data-driven product, these three…
