Mejdoubi M.

Mission: deployment of a GenAI strategic analysis assistant enabling Corporate Strategy teams to query a heterogeneous corpus (sector reports, market data, KPIs) in natural language.

Technical environment: Python · LangGraph · OpenAI · Milvus · RAG

Spanish multinational group, operating in 13 African countries, revenue: approximately 1 billion USD, leadership position (N+2 to Group CEO)

Design and deployment of a cloud CDP on Azure Databricks: multi-source ingestion (ERP, CRM, IoT, field), Medallion architecture (Bronze/Silver/Gold), 360° Customer views. Definition of data model and governance (quality, ownership, data dictionary).

- Multi-criteria retailer segmentation: transactional scoring + geographic potential + socio-economic variables → sales force reallocation, visit prioritization, assortment optimization

- Clustering of active points of sale for range recommendations (assortment, refrigerators)

- High-traffic zone scoring: internal performance + external potential

- Dynamic RFM segmentation with temporal monitoring of segments

End-to-end architecture on IoT sensor data: ingestion, KPI calculation (opening rate, temperature, planogram), CDP integration. Automated alerting system by anomaly detection → proactive field interventions, improvement of fleet profitability.

Stream Lead on the implementation of a centralized commercial intelligence platform (UVE). SME on AI recommendation functions & suggested orders integrated into the digital ordering app (Super App).

Management of Market Share PowerBI dashboards for the multi-country Management Committee (Morocco, Algeria, Ghana): competitive benchmarks, Nielsen KPIs, commercial performance.

Data Science Manager & Technical Lead| Oct 2019 – Dec 2023

Mission: create the internal Data Science function within the customer knowledge department (pricing, geomarketing, market research, data science) & industrialize ML use cases with high Marketing & Commercial impact within the big data stack (Cloudera Big Data platform).

Recruitment and skill development of an internal Data Science team from scratch.

Stream Lead on the delivery of 6 ML use cases in production (mixed internal + external team). Contribution to platform architecture: batch & streaming ingestion, unified data model.

- **Churn Prediction (Prepaid)**: gradient boosting on CDR features, monthly scoring ~10M clients → integration into CRM retention campaigns

- **Customer Lifetime Value**: predictive CLV model → differentiated treatment (service, network, promo)

- **Social Network Analysis**: modeling of customer interconnections (GraphX), influencer detection, contagion-based churn prevention

- **Propensity Models**: multi-product appetite scoring (data, family bundle, upgrade) → optimization of outbound campaigns

- **Lifecycle ML-driven**: automated acquisition / activation / retention / win-back sequencing driven by ML scores