Your ad hands one person the whole AI roadmap: marketing, content, CRM, loyalty, product.
Your Group CEO promised AI across the customer journey in December, and with a listing targeted for early 2027, every pilot now has to end in a number rather than a demo. So here is the week-one deliverable: the first slate, prioritized, each pilot carrying the number it must produce and the line under which I kill it.
Content velocityPilot 01 · weeks 1 to 6
Generative AI drafts product and campaign copy for 2,000+ celebrity boutiques and 1,000+ brands, English and Arabic together, and nothing publishes without an editor’s approval.
Number: days from brief to live in all 8 markets · Kill line: editors rate the drafts below parity in week 6
Win-backPilot 02 · weeks 2 to 8
Churn scores pick the lapsed shoppers most likely to return, and win-back journeys run through the channels you already operate: email, SMS, push, WhatsApp, in-app.
Number: win-back rate against an untouched control group · Kill line: under 1.5 points, stress-tested below
Campaign copilotPilot 03 · weeks 4 to 10
One automated weekly performance readout across your Kuwait, Dubai and Cairo marketing teams, with reallocation suggestions a marketer can accept or reject in a minute.
Number: conversion lift on the same budget · Kill line: suggestions nobody acts on by week 6
Virtual try-onParked, deliberately
The flashiest line in your own ad is the one I would not start: it needs shade-accurate 3D product data that does not exist yet, and it cannot produce a defensible number inside the window that matters.
Revisit once the content pilot has cleaned the product data it depends on
Pilot 02, stress-tested: the kill line is built in
Worked example, illustrative figures. No inside data.
- Win-back rate
- 2.0%
- Shoppers won back
- 2,400
- Recovered per quarter
- 44,400 KWD
Iterate six more weeks.
Real signal, not yet a scale case: tighten the segments and the creative against the same control group, then decide again.
Assumes 120,000 reachable lapsed shoppers and an 18.5 KWD average basket, replaced with your real numbers in week one. Kill line at 1.5 points, scale from 3.
Week one ends with this slate in front of leadership: every pilot with an owner, a number and a kill line. Week two starts the first one.
I built this page and the model on it myself, with the AI toolchain I would bring to Boutiqaat: Claude Code for the build, n8n for the agent automations behind my own products.
Why this role, why me
01The role is my current day job
Your first bullet asks for an AI innovation roadmap aligned with revenue, retention and efficiency.
At Swiss Post I turn AI opportunities into exactly that: a prioritized roadmap with KPIs, build-vs-buy calls, and delivery from concept to launch.
02E-commerce numbers, not slides
Campaign optimization and A/B testing only count when they move a real shop.
At Ifolor I owned the e-commerce ecosystem of a CHF 100M+ business, reporting to C-level: +9% conversion and +15% checkout step rate through research, testing and analytics.
03I build with AI myself
You want hands-on Generative AI, LLM and automation judgment, not vendor slideware.
Pedal Peak, my cycling platform, is a live product I ship end to end with LLM workflows and n8n agents; my studio smedium does the same for client work.
04The honest read: your Oracle screener
I have never run Oracle Cloud, and my platform depth is builder-grade: LLM APIs, n8n, Cloudflare.
Your own ad pairs this role with Data, BI and Engineering for infrastructure. What I bring is the half that is harder to hire: turning AI into piloted, measured business results.
The record behind it
- since Jan 2026
- AI Project Lead, Business Development · Swiss Post, Advertising, Zurich. AI business models from opportunity sizing to prioritized roadmap, build-vs-buy, and launch.
- Aug to Dec 2025
- In between jobs and own ventures. Built smedium to first clients with AI-automated workflows, grew Pedal Peak, went deep on LLM tooling; cycled through Togo and Benin.
- Oct 2024 to Jul 2025
- Senior Product Manager, Lead E-Commerce · Ifolor Group, Zurich. Owned e-commerce ecosystem and strategy of a CHF 100M+ business; +9% conversion, +15% checkout step rate.
- Jun 2023 to Sep 2024
- Lead Project Manager · Brixel, Zurich. The bridge between senior stakeholders at UBS and Baloise and the internal product team.
- Mar 2020 to May 2023
- Marketing & Growth Lead, founding team · WePractice (Migros Group). Two funding rounds closed; grew to 10 locations, 23 people, 170+ customers; 1,000+ client matches in year one.
- Sep 2019 to Sep 2022
- Growth & Venture Builder · Sparrow Ventures, Zurich. Go-to-market for several internal startups, validation to scale-up, experimentation to lift conversion and cut CAC.
- Jan 2017 to Aug 2019
- Intrapreneur, Innovation · Die Mobiliar, Bern. Three market pilots from MVP to launch (Smide, XpertCheck, Lizzy); market experiments that steered product and marketing calls.
- 2005 to 2016
- Commercial and product-marketing foundation. FMCG product management at Domaco, Cruspi and Promena; accounting at Kuoni and AMAG; commercial apprenticeship at Bridgestone.
The full CV, including the early career in detail: download the PDF or read it in the browser.
My first 90 days
Days 1 to 14
Listen, then audit
The hubs first: marketing, CRM, product and data in Kuwait, Dubai and Cairo, plus an honest audit of what GA4, Firebase and the CRM channels can feed a model today.
Days 15 to 45
Slate v1, pilot one live
The slate lands in front of leadership with numbers attached, and the content pilot goes live first because its data already exists.
Days 46 to 90
First verdicts
Week-6 kill-or-scale calls on the first pilots, a playbook written from what survived, and AI adoption training starting with the marketing teams.
About me
Away from work I am mostly on a bike. In 2025 I bikepacked through Togo and Benin, came home, and kept building: Pedal Peak for the cycling community, smedium for small businesses, both shipped with the AI tools I use every day.
My partner and I are planning our move to Dubai. I am Swiss, direct in the way that saves meetings, and happiest when a plan turns into something people actually use.