Simon Willison¶

Overview¶

Simon Willison is an independent software engineer and writer, best known as the co-creator of Datasette. He writes extensively about LLMs, coding agents, and the practical architecture of AI-assisted software development — making him one of the most grounded, empirically-focused voices in the AI engineering space.

His work on simon-w-agentic-engineering-patterns provides a detailed field guide for building reliable AI-assisted coding workflows, drawing directly from his experience building and shipping tools with LLMs in production.

Key Contributions¶

Datasette — an open-source tool for exploring and publishing structured data
LLM evaluation frameworks — practical approaches to measuring AI output quality
Agentic engineering patterns — systematized techniques for building agentic workflows
The Unreasonable Effectiveness of HTML — argues HTML outperforms Markdown as an AI output format for rich, interactive artifacts

Relevance to dark-factory-kb¶

Simon's agentic engineering patterns are foundational to the kelly-factory-overview. His work on HTML artifacts for AI output also directly informs the output quality improvements discussed in the factory's QA gates.

simon-willison-unreasonable-effectiveness-html — his piece on HTML as an AI output format
simon-w-agentic-engineering-patterns — broader agentic engineering field guide

Source¶

Primary sources: Simon Willison's blog at simonwillison.net and Substack at simonw.substack.com.

Bibliography¶