THE MANIFEST
Things I've shipped.
Ten years across payments, sales tech, hospitality and ads. What was broken, what I built, and what changed.
the honest version
Money in, friction out
The setup. Digistore24 moves about a billion dollars a year for people selling digital products. When I joined, every payment leaned on a patchwork of external processors with nothing in between: if one declined, the sale just died. I became the company’s first payments PM, which is a polite way of saying the plumbing became my problem.
What I built. The company’s first payment orchestration layer, routing every transaction by card, geography and risk, with automatic fallbacks when a processor says no. Then the rest of the stack: Apple Pay, Google Pay and Amazon Pay across the US and Europe, six new currencies, a rebuilt credit card flow replacing eleven-year-old code, and dispute automation that resolves chargebacks before a human has to look at them.
What changed. Failed payments now get a second chance instead of a funeral. Fallback routing alone has recovered over $6M in sales that would have quietly vanished. Wallets became some of the best-converting ways to pay, buyers pay in their own currency, and the ops team stopped triaging disputes by hand.
how a declined payment gets a second chance
A product category, absorbed
The setup. Outreach is a sales execution platform used by hundreds of thousands of sellers. I owned the records everything else depends on: prospects, accounts, opportunities. Meanwhile, reps were researching target accounts in Word docs and mapping org charts in standalone tools built for exactly that one job.
What I built. We folded a standalone forecasting product into the core platform: one login, inline editing, real-time sync to Salesforce and Microsoft Dynamics. I shipped AI Account Planning to general availability, and built Prospect Relationship Map on my own: who reports to whom, who actually decides, drawn from the emails, calls and meetings the platform already knew about.
What changed. Account research went from an afternoon of tab-hopping to generated in place; reps told us it saved around 20 hours per account. New customers named these features in over $1.5M of signed deals. And a whole category of standalone mapping tools quietly became a feature.
the plan up top, the map underneath
The pricing layer under 3,000 hotels
The setup. Mews is a property management system for 3,000+ hotels. I owned the pricing services API, the layer every room, breakfast and spa booking flows through, including for partners like Booking.com. Waiting on my desk: an enterprise chain had been contractually promised age-based pricing, it was overdue, and nobody had cracked how to build it.
What I built. Age-based pricing sounds like one field until you trace it through reservations, billing, distribution partners, per-country legal rules and revenue recognition. We mapped the blast radius, refactored the pricing core, and shipped it. Then dynamic pricing on top, so revenue managers could price against demand and competitor rates instead of a spreadsheet.
What changed. The overdue contract unblocked, and Mews could finally sell to family hotels and big chains that treat children’s pricing as table stakes. The features drove 50K+ bookings and over €30M in revenue across 200+ hotels.
one rate card, four systems that had to agree
Zero to €400K a month
The setup. My first product job. ROI Hunter helped e-commerce brands run social ads and wanted a second act. Google Shopping was the obvious bet, except the product didn’t exist and neither did a PM for it. That was the offer.
What I built. The Google Shopping product, from scratch, working directly with Google’s partner managers to unlock API capabilities as we went. And because early products don’t sell themselves, I ran go-to-market too: positioning, sales enablement, the material that closes deals.
What changed. The product scaled to over €400K in monthly managed ad spend and brought 50+ high street fashion brands onto the platform. It also set the lesson I’ve reused ever since: shipping the thing is half the job. The other half is making sure anyone cares.
Linen Overshirt
€49.00
HIGH ST. FASHION CO.
MANAGED SPEND, €0 TO €400K+ / MO
a shopping ad, assembled from a product feed
After hours
Built on nights and weekends with AI agents. Some lived, some very much didn't, and all of them taught me how these tools actually think.
vibe-coded, no regretsFamily Finance
A real-time dashboard for couples to track their combined net worth, without handing bank logins to a third party. The nicest thing I've built outside work, and it lives in the open.
SEE IT ON GITHUBAirial
A Flighty-style flight tracker for Android, born of pure aviation nerdery: PRD, mocks in Stitch, FlightAware data, built on Replit. It logged flights and plotted them on a map. The live tracking never took off, and my respect for the Flighty team only grew.
Motion Mate
My first AI-built app: a no-fuss workout planner with an exercise library by muscle group and the matching YouTube video playing right inside the app. Most apps that do this charge money or bury it under features nobody asked for.
In the hangar
More side projects on the way. Ideas currently outnumber weekends.