Decisioning and personalisation
Recommendation systems, CRM orchestration, audience strategy, customer relevance, and the product choices behind scalable personalisation.
/ about / enterprise AI product leadership
Robin Cartier
I build the decisioning and measurement systems that help large organisations decide what to personalise, automate, and trust.
Group Product Manager based in Paris. At Decathlon Digital, I lead CRM and personalisation platform work across 67 markets and 200M+ customers, connecting customer data, AI, experimentation, and product strategy into systems teams can trust and scale.
200M+customers reached by personalisation systems
67markets across Decathlon's global footprint
+110MEUR business impact associated with CRM capability rollout
~40product, engineering, and data teammates in scope
/ point of view /
Useful AI improves decisions.
The real work is not the demo. It is making customer data, model capability, business rules, measurement, and operating workflows fit together well enough that people can use them with confidence.
- Personalisation is infrastructure. It should help teams choose the right message, recommendation, audience, offer, and channel with evidence.
- AI needs measurement discipline. If a team cannot explain what changed, who benefited, and what trade-off was made, the system is not ready to scale.
- Agentic workflows need trust. Autonomy only matters when governance, feedback loops, and human judgement are designed into the operating model.
/ current work /
CRM and personalisation at enterprise scale.
At Decathlon Digital, my work sits between customer experience, CRM platforms, data products, experimentation, and market adoption.
I help teams turn customer data into clearer decisions: what to recommend, when to personalise, how to measure impact, and how to scale across markets.
DecathlonSkyComcastNOW TVMediaMathCapgemini
/ expertise /
Where I do my best work.
Measurement and experimentation
A/B testing, incrementality, business impact, KPI design, and the discipline needed to separate useful AI from expensive noise.
Agentic operating systems
AI workflows, knowledge graphs, source-backed research, tooling, and governance patterns that make teams faster without losing judgement.
Enterprise AIEcommerce AIRecommendation systemsCRM platformsAI product strategyGEO and AI search
/ selected proof /
AI for CRMRunning practical AI experiments with customer teams.Proof that the AI story is connected to operating teams, not only strategy decks.
PersonalisationKeynote on atomic content personalisation.How modular content and decisioning logic can make customer experiences more relevant.
MeasurementBehavioural tracking for streaming products.Data products and measurement systems for customer journeys at media scale.
AwardAdobe Experience Makers winner.Recognition for audience-driven customer experience work at NOW TV.
Knowledge graphA public AI research graph.Structured notes on production AI, agentic systems, recommendation, measurement, and AI search.
CodeProjects and working systems on GitHub.Technical proof around data science, recommendation systems, and AI-assisted workflows.
A few places where the work becomes visible.
AI for CRMRunning practical AI experiments with customer teams.Proof that the AI story is connected to operating teams, not only strategy decks.
PersonalisationKeynote on atomic content personalisation.How modular content and decisioning logic can make customer experiences more relevant.
MeasurementBehavioural tracking for streaming products.Data products and measurement systems for customer journeys at media scale.
AwardAdobe Experience Makers winner.Recognition for audience-driven customer experience work at NOW TV.
Knowledge graphA public AI research graph.Structured notes on production AI, agentic systems, recommendation, measurement, and AI search.
CodeProjects and working systems on GitHub.Technical proof around data science, recommendation systems, and AI-assisted workflows.
/ De:Coded /
The AI signal, decoded for real decisions.
I write De:Coded to turn the latest AI news into practical judgement for people building, buying, measuring, or governing AI systems.
/ social proof /
How colleagues describe working with me.
Selected recommendations, lightly edited for length.
Robin's structured way of working creates clarity and momentum, which makes collaboration both effective and enjoyable.
Robin repeatedly showed the value of strong stakeholder management, considered strategy, and responsible use of data at scale.
You can rely on his ability to analyse problems thoroughly and come up with practical, pragmatic solutions.
/ credentials and recognition /
Still learning, still shipping.
Udacity Agentic AI NanodegreeAgentic AI, autonomous workflows, and applied tooling.
MIT CSAIL: Generative AI strategyBusiness implications and strategic use of generative AI.
Grenoble Ecole de ManagementMSc Management and advanced internet strategy.
Adobe Experience MakersBest audience-driven experience for NOW TV.
/ connect /
Follow the work, or start a conversation.
Best route for professional context is LinkedIn. Technical proof lives on GitHub. Direct notes can go to hello@robincartier.com.