Scoped to your situation.
Let's start with a 30-minute discovery call, and scope and fee are agreed in writing before any work begins.
01 · Advisory area
AI-empowered development
AI has changed what a product team can ship. Most development processes still assume the old bottleneck.
Typical questions
- How should discovery, validation and delivery change when a working prototype takes an afternoon?
- Which ceremonies still earn their place, and which are habit?
- What governance keeps an AI-fast team safe instead of noisy?
02 · Advisory area
Enterprise data platforms
The hard question has moved up a layer. Storage, compute and the modern data stack, even catalogs are close to be solved. What is unsettled yet is whether a machine really knows what the data and the process means and it be trusted to act on it. A human analyst absorbs that ambiguity without noticing; an agent cannot.
Typical questions
- When two systems disagree on what "revenue" or an "active customer" means, whose definition does an agent get to use, and who owns that call?
- Do we need already an enterprise context layer of catalog, semantic model, knowledge graph, policies, lineage and agent-facing tools, and what is the smallest version that earns its place?
- Which domains deserve real data products with named owners, contracts and quality guarantees, and which are fine as they are?
- What may an agent read, infer and act on by itself, how is that enforced at runtime, and how do we know afterwards what it did?
- Where do we hold out for open contracts between the layers, MCP, Open Semantic Interchange, data contracts, instead of buying one more vendor's proprietary context silo?
03 · Advisory area
Geospatial + AI
My road in the geospatial world (GIS) has brought from digital maps to telecom positioning systems, from 3D offline maps in smartphones to real-time freight visibility in the European leading FTL transport platform. Now AI has changed the analysis. Spatial mistakes there do not announce themselves: degrees used as metres return a plausible number, a broken polygon topology quietly drops the features that cross it, and the map still renders. When a machine is doing the work, the guardrails and the evidence matter even more as the analysis itself.
Typical questions
- Where can spatial analysis or AI move a real number: routing, ETAs, utilization, siting, network planning, and what is the evidence it pays back?
- When an agent hands us a spatial answer, what tells us it is right rather than merely plausible?
- Could we reproduce a location analysis a year from now for a regulator, a customer or a court, with sources pinned, assumptions written down and a clean rerun that does not depend on anyone's chat history?
- Do we know the licence and provenance of every location dataset we have built on, and what share-alike terms oblige us to do when we redistribute?
- Which geo capabilities are worth owning, and where is a hosted service for basemaps, routing, geocoding or elevation simply better than what we would run ourselves?
04 · Advisory area
Logistics technology
Marketplaces, visibility platforms and data products for transport and supply chain, from someone who has led product in all three.
Typical questions
- Which logistics data could become a product customers pay for?
- How do we turn visibility data into retention, margin or a new offer?
- What should our platform integrate next, and what should it stop doing?
05 · Advisory area
AI opportunities using company brains
Most companies have by now way more AI ideas than evidence. I help find the few worth funding, and structure organizational knowledge so AI can use it.
Typical questions
- Which of our AI or data ideas are real, which are premature, and which are not AI at all?
- How do we size the value of an opportunity before building anything?
- What does a company brain look like in practice, and where would it pay back first?
Ways of working
How an engagement takes shape.
- 01Working sessions and advisory days with your leadership or product team
- 02Independent assessments: a clear, evidence-based verdict on a decision in front of you
- 03Fractional product, data or AI leadership, part-time, for companies scaling without a full-time CPO or CDO
Advisory only, and vendor-neutral. I do not sell development capacity, and scope and fee are always agreed in writing before we start.
See a sample AI-driven analysisNot sure where to start?
Most conversations start before the label is clear. A free 30-minute discovery call is enough to see whether I can help, and how.