Strategist and cultural researcher. I prototype with code to find out what’s worth building and how it will land with real people.
One claim · three ways it works · three projects that prove it
Models, avatars, and pipelines are now commodities. What decides whether an AI product survives is whether people adopt it, trust it, and fold it into their lives — which low-stakes uses they try first, which rituals form around it, when a device stops being equipment and becomes identity.
That arc is a cultural question before it is a technical one. My research treats frontier technologies — agents, digital twins, neural interfaces — as one story: agents act on your behalf, a digital twin is the model of you they act from, and sensors are the signal feeding that model. The product question of the next decade is what computing becomes when it responds to your state instead of your commands, and whether people will trust it.
I don’t research the frontier from a distance. For the neural-interface work, I wore the sensors myself.
What should AI-powered computing become over the next decade, and how do frontier technologies like digital twins and neural interfaces earn a place in everyday life?
Co-developed a 10-year, society-centric vision through a cultural research lens rather than a capability roadmap, delivered at steering-committee level.
The vision now informs how the lab’s design leadership frames its long-term product direction.
A hardware-first company wanted to enter the outdoor BBQ category. Could software lead instead of another device?
Led category entry strategy between a Berlin design agency and the client’s New Business team: hypothesis-driven research, eight in-depth user interviews, and a working AI companion concept structured around planning, cooking, and learning.
All eight users validated the planning value; the board got a decision framework for digital-first category entry, shifting the conversation from “what does it cost” to “what are we answering.”
Parliamentary inquiries were processed through fragmented systems and person-dependent knowledge, with context lost at every handoff.
Conducted discovery research, designed a local-first architecture, and — as a non-developer — built a working prototype with AI tools: semantic search over historical inquiries, running entirely on government infrastructure with no cloud dependencies.
A functional prototype and implementation blueprint, open-source-ready for replication by other municipalities, now informing the RFP for production development.
A newly formed innovation unit needed strategic direction on emerging technology, and the internal legitimacy to act on it.
Led a seven-week foresight sprint producing four 2035 scenarios, written as fictional news articles with a science-fiction author: grounded enough to be credible, uncomfortable enough to provoke change.
The modular report became the unit’s change-management instrument across the organization; the pandemic accelerated several identified trends within months, giving the scenarios immediate credibility.
Most of my projects are not about AI at all. They are about contact lenses, cars, drinks — categories with cultural problems. But the way I work is AI-saturated: more data in, faster to the point, more room for the creative work that actually moves a strategy.
I run my own research infrastructure, and I’ve written publicly about what delegating to machines forces you to articulate about your own method. The result is cultural research with unusual breadth that still ends in frameworks a board can act on — like the Visibility-Shift Matrix, built to show how culturally invisible products gain identity meaning.
Rising need among young consumers, falling category entry: the product was invisible twice — physically (worn unseen) and culturally (no identity meaning) — while glasses became a Gen Z identity kit.
Diagnosed the cultural barrier through ethnographic research and built the Visibility-Shift Framework from seven case studies of brands that made invisible categories culturally expressive.
The framework became the backbone of the category expansion strategy, and reframed M&A from growth tactic to cultural capability acquisition, informing year one of the three-year growth plan.
An iconic brand needed differentiated relevance in the luxury wellness space, where the category was being redefined beyond comfort-and-technology codes.
Led strategy in a six-week sprint: cultural wave mapping with anthropologist Grant McCracken, twelve target-customer interviews across five countries, and competitive analysis spanning legacy luxury, EV challengers, and wellness-first Chinese brands — synthesized into three strategic brand territories.
The framework became the foundational architecture for the brand’s global marketing, partnerships, and product innovation decisions.
Alcohol consumption is declining across categories while Gen Z holds multiple contradictory mindsets at once. How does a beverage portfolio stay relevant?
Served as senior cultural advisor to a future-of-drinking-culture program: enriched a six-mindset framework with international signals (soft clubbing, third spaces, little-treat culture) and challenged it to include escapism as a missing dimension.
Shaped the final recommendations toward multi-brand strategy and occasion-based innovation.
Most organizations treat strategy as a story: an offsite, a deck, a memo. AI exposes the weakness, because you cannot delegate what you cannot articulate. Strategy has to become explicit — readable by people, executable by machines — so decentralized teams can improvise within a shared groove.
The objection is always the same: won’t that kill creativity? Fela Kuti answers it. Afrobeat’s hour-long improvisations weren’t freedom from structure; they were built on extraordinarily tight grooves, internalized until improvisation could happen within them. Structure, collectively internalized, is what makes the emergent magic possible.
In early 2021, AI-generated content was either sensationalized or dismissed — and a public broadcaster needed an actual position on what it meant for media production.
Led research and strategy for the first futures report by any media company on AI-generated media: four scenarios for 2023–2026, expert interviews, and a voice-clone prototype of a popular radio host that made the implications visceral for management.
Published May 2021, 18 months before ChatGPT. When generative AI went mainstream, the organization already had vocabulary, frameworks, and a documented position.
Everyone agreed the broadcaster had to prepare for disruption; nobody had institutionalized the capability. Should there be an innovation lab, and what should it be?
Designed the answer as an institution: a 100-person organizational event to map readiness, scenario work translated into opportunity spaces, and a positioning model close enough to operations to matter yet protected enough to keep a long-term view.
The Hub survived its three-year pilot, became permanent, grew to ten people, and produced the Synthetic Media Report. Five years on, the relationship continues.
A B2B toy manufacturer wanted a direct-to-consumer presence, but D2C demands a content apparatus a small company can’t conventionally staff.
Developed an AI-forward brand and content strategy: systems that let a small team produce customer-centric content at D2C scale.
The strategy enabled the company to establish a D2C presence beyond its pure-B2B model — an opportunity previously out of reach at their size.
Advisor & Strategist · Berlin · Everything to someone, not something to everyone