Practical writing on data, AI, and impact

Dashboard or signal? The question under the reporting question

Boards often ask for a better dashboard, while the real split is signalling versus reporting. One shows what happened; the other acts when it counts.

Customer portals without manual work: from 3 to 30 clients

A portal that works at three clients often stalls at thirty. How row-level security and a multi-tenant setup replace manual rights work.

Row-level security explained for boards

One shared report, each user only sees the rows they are entitled to. How boards share data with staff, partners and customers without losing control.

Having data and understanding data are two different things

Storage keeps the data. Semantics records what it means. A data project that only looks at tables gets stuck on interpretation differences between departments.

Why an AI with access to your database still gives the wrong answer

Full database access is not enough. Without semantic context, an AI fills meaning gaps itself and still gives the wrong answer.

The data dictionary that gathers dust, and the definition-in-code that doesn't

A semantic layer ties technical fields to business definitions, so terms like revenue and active customer mean the same thing everywhere. Without it, an AI agent quietly keeps calculating on outdated rules.

Who gets to decide what "active customer" means?

Ask five people what an "active customer" is and you'll get five answers. Why ownership of data definitions isn't an HR question but a technical agreement.

"We've already connected everything" — four quiet signs that you haven't

Many boards think their systems are connected, while in practice separate links run that don't quite complete each other. Four quiet signs that "everything's connected" still doesn't add up.

From source to answer: what happens between your ERP and an AI answer

The full chain from source system to AI answer, explained in six stations. What happens at each one and what breaks if you skip it.

The six capabilities that make you BI- and AI-ready

AI-ready is not a certificate but an operational capability. Six questions with which you set your own starting position, with a test per capability you can run yourself.

Why AI projects fail on the data foundation

Why AI pilots fail on the data foundation, which four problems almost always sit under a failed pilot, and the sequence in which you actually build it up.

From Disconnected Data to a 24x7 Data Analyst

Always-On Reporting and Insight for You and Your Customers

When data lives across disconnected systems, reporting becomes slow, manual, and inconsistent. Impacture turns that fragmented landscape into an always-on intelligence layer with trusted dashboards, automated reports, and one shared version of the truth.