Independent AI consulting & implementation

AI that lands in your processes.

I help mid-sized businesses reduce document-heavy work – order intake, invoices and records, compliance – with AI. From opportunity assessment to a running system.

30 minutes, free · A direct conversation with me.

For operations and product leaders in mid-sized businesses. Whether it's orders, invoices and delivery notes, compliance/EHS records or customer service: we start with your work, not the model.

Four ways to move forward

A clear first step.
A tangible outcome.

From the first decision to everyday operation. Start where your team is.

01

2 weeks

Opportunity Sprint

Explore your workflows, assess feasibility, and decide where AI is worth using – and where it isn't.

You leave with A prioritized use-case map and a practical roadmap.

from €4,900 net

Book an intro call
02

4–6 weeks

Pilot

Build a focused pilot with your documents and clear success criteria before committing to a wider rollout.

You leave with A working pilot and evidence for the next decision.

Price after intro call

Book an intro call
03

Duration by scope

Integration

Connect the workflow to your tools and data, with appropriate access, human review, and reliable operation.

You leave with An integrated workflow with evaluation and clear ownership.

Price by scope

Book an intro call
04

1 day

Team Workshop

Practical work on your own use case, with documentation and hands-on handover. Help your team use the system confidently.

You leave with A team equipped to use, maintain, and improve the workflow.

from €1,900 net

Book an intro call

All prices are net, excluding VAT.

The person behind the work

I'm Fabian.

Consultant, developer, and your direct collaborator.

My background spans enterprise systems at SAP and Siemens, customer-facing products at Enpal, and business processes at HARTING.

You work directly with me – from understanding the problem to building the solution and handing it over. I connect the technical decisions with the people doing the work.

Best fit: businesses with 50–1,000 staff whose operations or product teams want to improve a document-heavy workflow. Other concrete projects are welcome, too.

Professional background

Experience with systems.
And their everyday use.

  1. 2024–2025

    SAP

    Backend development & GenAI workshops

    Product Compliance and Environment, Health & Safety: complex enterprise workflows with real operational constraints.

    My contribution: Built backend paths, explored GenAI use cases, and led workshops within the team.

  2. 2023

    Enpal

    Full-stack development

    A solar customer portal: helping customers navigate a complex product and service journey.

    My contribution: Enhanced the customer portal through full-stack development, making information easier to understand and act on.

  3. 2019–2023

    Siemens

    Mobile apps & dashboards

    Operational data for the team's day-to-day decisions.

    My contribution: Developed mobile applications and operational dashboards that made live data usable in everyday work.

  4. 2015–2023

    HARTING

    Data, web & ERP

    Software inside business processes across the web shop and operations.

    My contribution: Integrated data-science tools, web applications, and SAP ERP into those workflows.

Company names describe experience, not client endorsements.

Research & experiments 05

01 · Safe architectures

AI Mathematics

Constraints + correction paths

Small, falsifiable studies into systems with explicit constraints and correction paths. Research, not a claim of general AI safety.

Formal methodsResearch notes

02 · Learning systems

Learn When It Matters

Selective learning + measurement

A bounded investigation into selective learning updates and what it takes to measure them honestly.

MeasurementSelective learning

01 · Play · systems thinking

Scene Detective

Patterns + better questions

A compact game about noticing patterns, testing hypotheses, and learning to ask one better question.

Game designInference

02 · Creative tools · visuals

Phosphor

Responsive worlds + creative coding

A visual instrument for rehearsable, responsive worlds—an intersection of engineering and creative practice.

Creative codingInteraction

03 · Simulation · replay

Tenzo Lab Local

Incentives + fragile signals

A paper-only simulator for exploring incentives, lag, and fragile signals. No live trading or real-money results.

SimulationPaper only

How we'll work together

Start small.
Make it stick.

  1. Understand the work

    Start with the people, the workflow, and what's getting in the way. Agree on what a better outcome means.

  2. Test the smallest useful change

    Choose one use case. Check the data and risks, build a pilot, and compare it with the current way of working.

  3. Build for everyday use

    Connect tools and models where they earn their place. Add evaluation, clear permissions, and human review.

  4. Leave the team in control

    Train the people using it. Document decisions, hand over ownership, and agree on how to measure and improve it.

Questions before we start

What you should know.

How do you protect our data? Where does it live?

I use EU-hosted models by default, such as Azure OpenAI in EU regions or Mistral. When data must not leave your company, we use open models on your own servers. Your data is not used to train models. Every project includes a data processing agreement (AVV).

What does the EU AI Act mean for our project?

Most pilots fall into the minimal- or limited-risk categories. In the Opportunity Sprint I classify the use case and document the transparency obligations. This is not legal advice.

How long does a project take, and what does it cost?

The Opportunity Sprint takes 2 weeks and starts at €4,900 net. A pilot takes 4–6 weeks; the price is agreed after the intro call. Integration and operation are priced by scope. The Team Workshop takes 1 day and starts at €1,900 net.

Explore all packages
What do we need to provide?

A named owner, access to example documents or data (anonymised is fine), and 2–3 hours per week from the people doing the work during the sprint. We also need a decision-maker for the go/no-go decision.

Which models and tools do you use?

I'm vendor-neutral. Models and tools are chosen for the use case, your existing systems, and your data-protection requirements – not the other way around.

Your next step

Which workflow
could be easier?

Tell me what's slow, repetitive, or unnecessarily complicated. You don't need a finished AI brief to start a conversation.

30 minutes, free · A direct conversation with me.