Doug Turnbull softwaredoug.com

AI coding needs test train splits

softwaredoug.com·4w

Semantic Search Without Embeddings

softwaredoug.com·5w

Free course: Cheat at Search Essentials

softwaredoug.com·8w

RAG users want affordances, not vectors

softwaredoug.com·9w

LLM Judges aren’t the shortcut you think

softwaredoug.com·14w·

Discuss: Hacker News

An agent-coded search reranker

softwaredoug.com·16w

General purpose agentic loop in 40 lines of Python

softwaredoug.com·17w·

Discuss: Hacker News

Adding reasoning improves search relevance 15-30%

softwaredoug.com·18w

Agents turn simple keyword search into compelling search experiences

softwaredoug.com·20w

Reasoning Agents prefer Simple, Transparent Search

softwaredoug.com·20w

BM25F from scratch

softwaredoug.com·21w

BM25F from scratch

softwaredoug.com·21w

The scapegoats guide to transforming organizations

softwaredoug.com·21w

The scapegoats guide to organizational 'transformation'

softwaredoug.com·21w

Do well-written, clear instructions beat few-shotting for tiny-LLMs?

softwaredoug.com·23w

Few shotting vs Rules with Small LLMs

softwaredoug.com·23w·

Discuss: Hacker News

OpenAI lost the plot on 'boring' LLM use-cases

softwaredoug.com·25w

OpenAI lost the plot on 'boring' LLM use-cases

softwaredoug.com·25w

Researching with Agents

softwaredoug.com·25w·

Discuss: Hacker News

Free course: Cheat at Search: Essentials

softwaredoug.com·28w

Loading...Loading more...