I find where AI pays off, build it, and prove it works.
I'm Jan-David Stütz, an AI engineer with a PhD and nine years of shipping applied AI. I work inside your team and your codebase: I talk to the people doing the work, find where AI can take load off them, and bring you a concrete proposal. Then I own it, from concept through tests to daily releases in production. In user studies, my systems cut the time for specialist work by up to 80 percent, and one opens up an estimated €72 million a year in potential savings.
Some problems need a simple assistant, others several agents with memory. Either way, I test the system against prepared scenarios before anyone relies on it, and I mentor your team along the way so the know-how stays when I leave.
At a glance
9+ years
Shipping applied AI products
6 papers
Peer-reviewed evaluations of my systems
1 owner
Strategy to production, end to end. A vetted network joins when a project needs it.
Featured Projects
Selected case studies
Three recent projects, one from each shape of AI work I build.
Impact
What the work delivers
The through-line across my work is measurable business impact: interfaces that let non-experts deliver what used to require specialists, faster and at higher quality. Knowledge graphs get built 42 to 83 percent faster, and mapping quality rises from 54 to 92 percent compared with hand-written code [DB2KG, Text2KG]. People who can't write a query build their own dashboards, with usability rated excellent [DIVE]. Goal-driven recommendations cut shopping time by two thirds and mental workload by 41 percent [GoRS]. Tail-end negotiations run automatically, with an estimated €72 million a year in savings potential [Negotia], and a sustainability-aware procurement workspace runs inside a global Fortune 500 company [DPW]. Every result comes from a controlled user study, and all of these systems run in production today in a further developed form.
Who I work with
How to start working together
01
I have a problem, not a solution
Workflow diagnostic
The diagnostic start. I study the workflow, talk to the people in it, and come back with a map of where automation pays, where humans stay, and what to build first.
02
I have a direction, need the right build
End-to-end delivery
I join your team and work in your codebase. I find where AI pays, propose it, and own each feature from concept to daily releases in production, tested and measured. No handoff chain.
03
I have a system, need it better
Improve a live system
Iteration on something already live. Adoption is stuck, outcomes underperform, or the human-machine split needs rethinking. I evaluate, rebuild the weak joints, and measure again.
Let's find the work AI should be doing for your company.
You'll hear back within one business day, and the first conversation is a free 30-minute call.