Blog
Notes from the maze
Practical writing on the problems we help businesses solve: AI, machine learning, data and the rules that govern them.
EU AI Act readiness: where to start if your company uses AI
A practical first pass at the EU AI Act: inventory your AI systems, classify them by risk, and close the gaps before obligations apply.
Read article →AI harness engineering: what makes an AI agent dependable
The model is only part of an AI agent. Tools, memory, guardrails, evaluations and monitoring decide whether it does the job the same way every time.
Read article →Building an AI assistant on your own documents
How retrieval-augmented generation (RAG) lets an assistant answer from your company's documents, and what decides whether it gives correct answers.
Read article →Forecasting and pricing models that keep working after launch
What separates a machine-learning forecast that impresses in a notebook from one a business can rely on month after month.
Read article →