Decathlon
What Problem Did We Solve for Our Customer?
In today’s rapidly evolving retail landscape—where digital expectations meet sustainability imperatives—a leading global sports retailer set out to rethink how knowledge is discovered, shared, and actioned across the enterprise. With vast operations spanning supply chain, marketing, and customer interaction, fragmented data was limiting the organization’s ability to innovate at speed and scale.
The vision was bold: a unified intelligence layer that could seamlessly weave together business knowledge, contextual data, and operational insight. This would allow every team—from retail operations to logistics planners—to tap into a shared source of truth, enabling smarter decisions and faster innovation.
Scaling AI capabilities across the business without losing grip on direction, purpose, or pace was critical. There was a growing need to:
- Unify fragmented knowledge.
- Boost semantic search performance.
- Prepare for the next generation of customer experiences powered by AI but grounded in real business value.
The goal was to evolve AI capabilities from siloed experiments to a business-critical, scalable capability embedded across teams, tools, and touchpoints—moving from traditional IT to truly competitive digital infrastructure.
What Did We Create/Develop to Solve This Problem?
More than just technical delivery, Lionsville designed and built the operating model that connects AI, IT, and the business—turning abstract potential into tangible competitive edge. In co-creation with the customer, we established a dedicated Knowledge Graph Engineering Team.
Our Approach
We began by deeply embedding with teams across business and technology domains. Through collaborative research and exploration, we uncovered the hidden landscape of enterprise knowledge: disconnected systems, untapped datasets, and siloed expertise. We then orchestrated the collection and activation of data across the organization, building a semantic Knowledge Graph platform that could interconnect people, processes, and products in meaningful ways.
Core Platform Features
At the heart of this platform, we developed a custom ontology tailored to the organization’s domain-specific context. This ontology enabled the platform to interpret customer, product, and operational data with high fidelity—empowering teams to:
- Extract meaning from their existing assets.
- Connect insights across silos.
- Maximize the value of existing data.
This foundation was not just about structure—it was built for performance and trust, delivering:
- Low-latency access.
- Robust security.
- Enterprise-grade scalability.
- Compliance with strict infrastructure and software policies.
- Governance principles ensuring responsible and transparent AI use.
Real-Time Use Cases
By integrating data from across retail, logistics, and customer touchpoints, we enabled real-time capabilities, such as:
- Understanding product intent in customer queries.
- Enriching catalog data with multidimensional insights.
- Layering advanced AI services to create an adaptive digital assistant capable of answering complex, business-grounded questions.
The Result
We delivered intelligent tooling that supports:
- More contextual product discovery.
- Anticipation of customer needs.
- Acceleration of digital initiatives across the company.
What began as an internal effort to better structure knowledge became a catalyst for strategic transformation—empowering teams to reimagine the product lifecycle, design resilient logistics flows, and elevate retail’s role in the digital ecosystem.
Strategic Layers of Work
- Business Awareness – Making the vision visible, aligning AI goals with real business outcomes, and translating complex architecture into clear value streams.
- Whole System Thinking – Connecting the AI strategy to the broader ecosystem—from product teams to IT to leadership—ensuring buy-in, adoption, and scalability.
- Ownership & Direction – Designing and delivering a modular knowledge graph architecture that supports semantic search, AI-driven insights, and future e-commerce personalization at scale.
Together, we laid the data foundation and operational model for a smarter, AI-driven product experience—tailored to every sport, every customer, every interaction.
Technologies Used
- Knowledge Graphs: Neo4j, RDF/OWL, SPARQL
- Semantic Search and contextual product tagging
- Python & LangChain for interactive reasoning
- Vector Databases & AI enrichment
- Azure Cloud Infrastructure
- Agile DevOps delivery model
🚀 Developed by Lionsville.