Consulting · JanJo Consulting

From concept to measurable impact.

I publish peer-reviewed research on whether people actually trust AI systems — and I ship those systems into production at an industrial company. Consulting sits exactly where those two things meet.

I take on a small number of engagements at a time, alongside a full-time research role. That keeps the work focused — and it means an early conversation is the one I can actually act on.

Selected work

Three problems, and what changed.

01

From scattered AI ideas to a funded roadmap

Situation
A DAX-listed industrial manufacturer had AI ambition spread across many business units — but no shared way to judge which ideas deserved investment. Every unit made its case in its own format.
Approach
Designed and facilitated a Data & AI Opportunity Workshop framework: a repeatable, single-engagement format that turns a unit's strategic priorities into a prioritised roadmap, with value and feasibility assessed side by side.
Outcome
Proven across multiple business units as a repeatable route from broad AI ambition to a sequenced, decision-ready roadmap.
02

Reading user feedback at a scale humans can't

Situation
Product teams held large volumes of unstructured user feedback — forums, tickets, surveys — that nobody could read at scale. Insight arrived too late to shape releases.
Approach
Architected an end-to-end LLM pipeline that turns raw feedback into structured, actionable insight — with domain expertise codified into the system rather than bolted on afterwards.
Outcome
Analysis that had taken hours of manual work now takes minutes. A European patent application is pending on the approach.
03

Turning a pocket of practice into a company-wide capability

Situation
Data-driven UX lived in isolated pockets — a handful of practitioners, no shared standards, and no obvious route for a team that needed help.
Approach
Founded and ran a company-wide Community of Practice: set the cadence, established shared methods, and connected teams to the people who could actually help them.
Outcome
Grew from 7 to 150+ members within twelve months, and became the default internal entry point for data-driven UX.

Engagements are described in anonymised form. These illustrate approach and capability — client intellectual property and confidential detail are not transferable.

Formats

Four ways to start.

01

Data & AI Opportunity Workshop

A facilitated session that takes your strategic priorities and leaves you with a prioritised AI roadmap — value and feasibility assessed side by side, and a clear view of what to fund first. One engagement, one decision-ready output.

02

Proof-of-Concept Sprint

A time-boxed build that answers one question: does this actually work with your data, your constraints, your users? It ends with a working prototype and an honest path to production — or an honest recommendation to stop.

03

Advisory Retainer

A regular cadence rather than a project. Architecture and roadmap review, a second opinion before the expensive commitments, and someone to think out loud with in between.

04

UX Research for AI

A multi-method study that checks whether the AI you are building solves the problem your users actually have — before the build cost is sunk. The same methods as the peer-reviewed work, pointed at your product.

Scope

Areas I work in.

  • Agentic AI & LLM systems End-to-end GenAI pipelines and AI agents, from foundation-model architecture through to production.
  • AI & data strategy Frameworks that survive contact with reality — designed across multiple business units, adapted to your context.
  • Trust & explainable AI Industry-oriented design guidelines for AI that users actually trust, informed by peer-reviewed research.
  • Affective computing Digital empathy and emotion-aware design, where measuring how something feels matters as much as whether it works.
How we'd work

A simple shape.

1 · Conversation. Tell me what you're trying to do and what's blocking it.

2 · Diagnosis. A short, written read of the problem and the cleanest path forward.

3 · Engagement. Either an advisory cadence or a focused build sprint, depending on what you need.

Start a conversation