Real customers try your product and prototypes, and AI synthesizes what they say and do. The synthesis feeds a knowledge hub, and your team's customer understanding grows with it. We develop this loop and its playbook together with our clients, pairing agentic tools with coaching. We call it agentic process co-design and run it inside product teams, including Bolt, mobile.de and Entirely.
Eight papers on designing an LLM wiki for customer research: what they tested, what we would build and what we have learned in practice.
Read the noteWhat we're learning at Bolt about building an AI user research agent on an LLM wiki, and about making shared knowledge work for a team.
Read the noteOn AI in customer research, continuous discovery and the work of keeping shared knowledge useful.
Read the noteTwo workshops, one pattern: non-technical teams built working LLM wikis with Claude Code in ninety minutes. The entry barrier has fallen. What matters now is designing the knowledge system the agent works in, not writing a better prompt.
Read the noteAlso in 2026
It automates the recruiting workflow for 100 interviews a week.
Read the caseProduct teams base decisions on what customers write.
Read the caseBased on real customer evidence, the leadership team settled on a clear business model.
Read the caseYou start with a 30-minute call. When it fits, we run a pilot with your real data. From there we work inside your team week by week: coaching, tools composed from what you already have, and workshops. Your team learns to use agents in its own product discovery.
How we work with you
Satya Nadella describes a learning loop in which human capital and token capital compound. We apply the idea to product discovery. The two sides are your team's judgment and the agents you own.
“You can offload a task, or even a job, but you can never offload your learning.”
Teams we have worked with
Bastian Gerhard
Managing Director, Entirely
“Based in Berlin, Stefan is one of the world's best innovation coaches out there, and my go-to-person for agile leadership, customer centricity and product development — not only for the tools and methodologies, but especially for the mindset.”
Markus (Yong-Min) Jo
Senior User Experience Researcher, mobile.de
“podojo transformed our approach to customer feedback analysis and helped make evidence-based product strategy decisions, automating categorization and saving us countless hours of manual work.”
Marek Janetzke
CEO, Lucid Labs
“Stefan has a knack for bringing clarity with the right questions. We worked on product topics together and I benefited from his deep experience — a solid methodological foundation combined with a contemporary understanding of AI in product management.”
Christian Marx
Zamics
“AI tools have revolutionized Product Discovery! During the AI Product Dojo, I learned how to automate time-consuming processes like creating user profiles or summarizing insights, while balancing the strengths and limitations of AI for evidence-based decision-making.”
Uwe Günzel
Co-Founder, RoX Health
“An impressive development within just six months. Our employee is noticeably more self-confident, performance and reliability have increased — and the whole team benefits.”
