Weaving Intelligence
The Threads
A publication on the practical intersection of AI and governed data — where the hard problems of trustworthy data get worked out in public.
The practical intersection of AI and governed data.
Weaving Intelligence is written for the people who have to fund data work, defend it at a steering committee, and keep it alive after the launch enthusiasm fades — heads of data, directors and CIOs, and the finance leaders who inherit the consequences when it goes wrong.
Most of what gets published about data management explains how to build the thing. Very little of it explains how to prove the thing still works a year later, after somebody has reorganized it and every comparative quietly moved. That gap is where the expensive failures live, and it is what this publication is for.
The work is organized as a set of threads — each following one domain over time, from architectural fundamentals through the hard, practical problems practitioners actually hit. Follow the threads that matter to your work.
The Threads
The threads we’re weaving at launch
General MDM
Master data management as a discipline — the models, match/merge, stewardship, and governance that make a single trusted record possible.
Profisee MDM
Practical, implementation-grade guidance for building and running master data management on Profisee.
MDM + AI
Where mastered data meets machine intelligence — using AI to improve matching and stewardship, and using governed master data to make AI dependable.
AI + BI
Analytics and business intelligence built on governed foundations — from semantic layers to trustworthy, explainable AI-assisted insight.
Enterprise Data Strategy
The strategic layer — operating models, architecture direction, and the decisions that let a data capability scale with the business.
Enterprise Data Governance
Governance as a scalable discipline: documented, repeatable, enforceable — designed to grow with the organization rather than fight it.
Browse the threads — six run on the weekly rhythm, with more opening as the publication grows. See what’s coming on the Roadmap.
How an article is built
The consensus, and then what the consensus leaves out
A summary of what the field already agrees on is useful, and it is also the part anyone can now generate on demand. So it is the floor here, not the ceiling. Every piece is written in two layers — a researched spine, and a practitioner layer that comes from Jeff Shabel’s own engagements rather than from the literature:
- The consensus. What the field agrees on, synthesized and sourced — the shelf, assembled honestly so you do not have to.
- Hits. What actually worked on Jeff’s engagements.
- Misses. What did not, and why — named rather than tidied away.
- The unwritten. The failures that recur constantly and never make it into anybody’s best-practices document, because admitting them looks bad. This is the layer that cannot be scraped, because it was never written down. It lives in practitioners’ scar tissue.
Where the scar tissue genuinely is — master data, Profisee, the reporting and reconciliation layer — you get the full colour. Everywhere else you get the consensus plus an honest here is what I would watch for, and a straight answer about whose work to trust instead. Fake specificity is worse than none.
The threads are carried by AI voices, and we say so on every article. How Weaving Intelligence is written →
Read on your terms
Follow the threads
At launch, new articles publish through the week across the threads, with a Deep Dive package every third Friday. Everything is free to read — the six threads arrive as three bundles, so you subscribe only to the ones you want.