Finalist
3 weeks · Carrefour · Bachelor 2
Beer Assortment Optimization with AI
Optimizing beer assortment across 3 French regions with an AI-powered Streamlit dashboard, a natural language SQL chatbot, and recommendation algorithms for store managers.
Business Context
Carrefour needed to optimize its beer category assortment across Grand Est, Normandie, and PACA regions. Consumer preferences vary significantly by geography, and store managers lacked accessible data tools.
Strategic Problem
How to optimize the beer assortment in Carrefour stores across three distinct French regions to maximize sales while giving store managers AI-powered tools to understand and act on their data?
Data Sources
Store-level beer sales data, regional consumer preferences, SKU performance metrics, shelf space allocation, and demographic data per region.
Methodology
Built a Streamlit dashboard with regional statistics visualization and AI-powered assortment recommendations for store managers. Created a chatbot trained on store data for conversational data exploration. Implemented natural language to SQL queries (e.g., 'which is the best-selling beer?') using AI to generate and execute SQL queries — giving non-technical store managers instant access to insights.
Key Results
Reached finalist stage. Delivered a complete data product: Streamlit dashboard with AI recommendations, conversational chatbot, and natural-language SQL interface for store managers.
Business Impact
Demonstrated how AI can democratize data access in retail — making complex analytics available to non-technical users through natural language interfaces.