The Blog
Practical writing on production AI, data architecture, engineering patterns and the decisions that determine whether systems actually work.
Why Your AI Pilot Won’t Reach Production
Why successful AI demonstrations so often fail to become reliable production systems — and what organisations should do differently before committing to the next build.
READ ARTICLE →
How a RAG System Leaks a Salary Review
How an apparently successful RAG system can expose restricted internal documents — and why permission filtering belongs before retrieval.
READ ARTICLE →
Your Agent Framework Is a Black Box. A DAG Isn’t.
Why production agent workflows benefit from named tasks, independent retries, durable approval gates and execution paths you can actually inspect.
READ ARTICLE →
Why We Publish Our Engineering Patterns
Why we publish the architectures, tradeoffs and failure conditions behind our engineering work instead of asking buyers to trust methodology claims.
READ ARTICLE →
When Not to Use AI
Seven situations where a rules engine, a report, a redesigned process or a human decision is a better investment than adding artificial intelligence.
READ ARTICLE →