Alexander
Witteveen.

• 01Elevated to Co-Deputy Global CTO (Apr–Dec 2024), reporting to the CDO, aligning 2,500+ engineers across the entire digital estate.
• 02Led the team that architected the AI-driven personalisation platform generating €170M incremental revenue (150% ROI) — 225M+ customers, 80 countries.
• 03Led migration to composable, event-driven architecture across 400+ touchpoints (IoT, mobile, POS, e-commerce).
• 04Rolled out unified global Membership Platform (120M customers) and mission-critical CIAM for 225M+ users.
• 05Drove digital revenue contribution from 17.4% → 20% of group through platform-led innovation.

• 01Scaled Technology Hub from 32 → 220+ FTE (316% hypergrowth in 18 months); 26% of hires from Google, Booking, Amazon, Philips.
• 02Pivoted portfolio toward AI/ML-integrated connected platforms — modernised Scopus (€450M) and Engineering Village (€60M+).
• 03Embedded Data Science into the Research Technology operating model via SVPG methodology.
• 04Built award-finalist onboarding programme for 2,000+ annual hires — eNPS +91, €13M+ saved in unproductive labour.
• 05Lifted engineering eNPS from −35 to +12 in 8 months; 3% non-regretted attrition after 18 months.

• 01Led digital transformation of pan-European e-commerce + retail across 35+ products, 10 countries, 60+ stores.
• 02Defined cloud migration strategy (Azure / Private Cloud) with API-led architecture (Mulesoft).
• 03Managed €10M+ technology budget; built and coached multiple DevOps teams across geographies.

• 01Reduced time-to-market by 21% across 20+ parallel SCRUM teams.
• 02Owned the CRA system processing €2B in annual transactions.

• 01Delivered for Sonos, KPN, T-Mobile, Rabobank, TomTom. 37% productivity improvement.

• 01Designed and executed €25M of programmes focused on AI and human-machine interaction — fifteen years before the current cycle.

• 01Led UX research and usability projects across Google's European product surface.

I speak on enterprise AI adoption, composable commerce and event-driven architecture at scale, building 200+ engineering hubs (the Elsevier 32→220 story), and how a CTO earns a P&L seat at the leadership table. Past stages include the DevOn Conference. For booking, send the event date, audience profile, format (keynote, panel, fireside), and topic angle via LinkedIn. Press portrait, short and medium bios, and a wordmark are available on the press section of this site; I confirm or decline within 48 hours for serious enquiries.

Yes — non-executive directorships, technology committees and advisory board roles are work I take on in parallel to a full-time executive role. The fit I look for: a business where technology, product or AI decisions are board-level matters (digital transformation, platform investment, M&A diligence, AI strategy), and a chair or CEO who wants real challenge in the room rather than ceremonial presence. Two to four hours a month is the typical commitment, plus quarterly meetings. I sit on advisory boards for two stealth companies today. The fastest route in is a short note via LinkedIn — the company, the stage, and the specific gap on your board.

Internal AI is a tooling-and-process problem before it is a model problem. The wins compound when three things are sequenced correctly. First, instrument the work — see where humans actually spend time on tasks an AI can plausibly absorb, and where they spend time on things only humans should be doing. Second, deploy narrow, well-bounded copilots into the specific flows that surface, rather than buying horizontal "AI for everything" platforms and hoping value emerges. Third, measure the lift in cycle time, throughput and quality — not seat adoption. The failure mode I see most often is procurement leading the strategy: enterprise licences signed before anyone has redesigned the process the AI is supposed to accelerate. Productivity gains in the double digits are reachable; they require the operating model to move with the tooling.

Customer-facing AI is an architecture problem before it is a model problem. The platforms that ship value at scale share three traits: composable, event-driven foundations that let AI features plug into real customer flows without bespoke integration each time; a data estate that is observable end-to-end so personalisation, recommendation and search can be trusted by the business; and an engineering organisation structured for production AI — with the data, ML, platform and product disciplines wired to ship together, not in parallel. The €170M incremental-revenue personalisation platform at Decathlon — 225M customers, 80+ countries — works because those three were resolved in that order. The failure mode is almost always sequencing, not model capability. Most enterprises have access to the same models; what separates the ones that ship customer value is the foundation underneath.