An extraction pipeline that cannot show its evidence has not finished the job. In certification work, a value on a certificate is only as good as the trail connecting it to the source document. In finance, a payment decision is only as defensible as the invoice line behind it. Yet most document AI systems deliver … Read more
The most dangerous extraction error is the one that looks right. Modern document AI tools return clean, complete, schema-conformant output almost every time. Every field is present, every type is correct, and the record loads into the enterprise resource planning (ERP) system without complaint. Stakeholders see well-formed data and read it as accurate data. Our … Read more
Most intelligent document processing budgets are optimising the wrong half of the problem. Procurement teams compare extraction accuracy percentages, pilot teams demonstrate extraction quality, and vendors compete on extraction benchmarks. Yet the cost that decides whether a deployment pays for itself sits elsewhere: in the review queue, where people check what the models produced. Our … Read more
Across this series, we argued for a particular way of measuring AI readiness: evidence over enthusiasm, scores delivered as starting points, progress tracked as a trend rather than filed as an event, interviews built with the discipline of hiring science, and a rubric its own makers sit first. Arguments that long invite questions, and the … Read more
The question was about incentives. How does this organisation reward people for adopting AI? The answer was honest, thoughtful, and entirely in the future tense. We are still figuring that out. Under the rule, an answer in the future tense does not score. So it did not score. The finding went into the report, and … Read more
The interviewer asks about data governance, and the answer arrives instantly, polished by a dozen board meetings. “We take governance extremely seriously. It is a top-three priority this year. The interviewer nods, writes nothing, and asks a smaller question. “Tell me about the last time a data problem stopped an AI project. What happened next?” … Read more