I own the commercial outcome, not the roadmap.
30/mo to
10,000/day
Identity checks scaled at Onfido, 99.9% SLA
weeks to
hours
B2B course setup after platform rebuild at Babbel
transactional to
subscription
Business-model transition to profitability at Honeypot
−27pp
Trial cancellations in first months at NeoTaste
Ten years building and running product organisations for subscription and marketplace businesses — as first product hire, through acquisitions and rebuilds, to global accountability across eight markets. Engineer by training. I run AI-native teams by building the tooling myself.
Six companies, the same job: build the platform, the team, and the economic engine for the next order of magnitude.
2026 — now
Head of Product — first product leader
Subscription restaurant marketplace, a six-figure subscriber base across four countries. Hired by the CEO to build the product function from zero: the org, the operating model, the multi-year strategy — and an AI-native way of working.
2024 — 2025
Product Leader, B2B Live (contract)
Led product for the B2B live-learning business — group classes for enterprise clients. Reframed growth from content depth to experience reliability, then replaced per-course manual operations with a platform foundation: unified course object, central event system, spreadsheet-free workflows.
2020 — 2024
Head of Product
Europe's largest reverse recruiting platform. Owned the portfolio through acquisition and restructuring, and led the transition from a transactional model to subscription — the decision that took the company to profitability.
2018 — 2020
VP of Product
EdTech across eight markets. Replaced per-campus spreadsheets with one platform covering the whole student journey — admissions, payments and delinquency, class management, assessments, job outcomes — and gave the company its first cross-market analytics layer. White-labelled it for enterprise reskilling.
2017 — 2018
Product Lead — verification
Owned the trust layer behind partnerships with global brands. Launched a verification SDK adopted by 10+ of them.
2015 — 2017
Product Lead — first PM hire
AI identity verification, later acquired for $650M. Built the product function alongside the CPO and led the KYC/AML suite through hypergrowth in a regulated market — by redesigning how product, engineering, operations and compliance worked together.
Each one contrasts the road not taken, the experiment that decided it, and what the input change did to the output metric.
BabbelB2B platform
Everyone wanted more content. The data said the ceiling was operational: every course needed engineer-weeks of manual setup, batch-launched once a month. We rebuilt the operational core instead of the catalogue.
Instead of deeper content, we bet on experience reliability and a platform foundation.
setup from engineer-weeks to hours, launch any day of the month
Honeypotbusiness model
Growth work on a transactional model was pushing water uphill. The real lever was structural: move clients to subscription, redesign activation around it, and let retention compound.
Instead of more funnel optimisation, we bet on a business-model transition.
the decision that took the company to profitability
NeoTaste0 → 1 org
First product leader in a scale-up: instead of writing an AI policy, I built the LLM voice-of-customer pipeline and self-serve analytics myself — then made that the standard the team and company work to.
Instead of mandating AI adoption, we bet on leading it by example.
trial cancellations down 27pp in the first months
I build the internal tooling myself — it's how the standard gets set. A selection, each one in daily use.
End-to-end subscriber feedback pipeline: LLM quality gate, embedding-based clustering, and a daily insight briefing agent delivered in Slack.
LLM pipelineembeddingsSlack agent
Schematic recreation with synthetic data — not a product screenshot.
Natural-language KPI access for the whole company — anyone can ask the data a question and get a governed answer.
Snowflake Cortex
Schematic recreation with synthetic data — not a product screenshot.
Planning tool wired into Linear and Notion via MCP, keeping strategy and delivery in one view.
MCPLinearNotion
Schematic recreation with synthetic data — not a product screenshot.
German-learning PWA in daily use: Leitner spaced repetition, speech practice, offline-first.
Next.jsTypeScriptAnthropic APIClaude Code
Interactive demo with sample vocabulary — a simplified version of the real app's Leitner scheduling.
I don't delegate the future of work to a policy doc. I build it, use it, and expect my teams to work the way I do.
Commercial outcomes, not roadmaps: the strategy, the team, the operating model, and the P&L conversation behind them.
Bad news travels fastest. Stakeholders hear the risk from me before they see it in a dashboard.