David OUEDRAOGO

1. CAIP Certification - AI Instructor (Private Training) – Created and delivered the CAIP (Certified Artificial Intelligence Professional) course to a student from Saudi Arabia, in Paris. – This experience was an exceptional opportunity for me to refine my teaching methodologies and revisit many key concepts in machine learning. 2. Wild Code School — Generative AI & Prompt Engineering Mentor – Led hands-on workshops on prompt engineering and advanced OpenAI API usage. – Supported students through Q&A sessions, debugging and practical exercises. – Guided final project development and provided continuous feedback. – Integrated new GenAI practices through active technology monitoring. 3. ESIEA — Data Crafting & Data Mining Instructor – Taught a full 42-hour module to 30+ MSc students (Software & AI track). – Delivered lectures on data quality (GIGO), data validation, and cleaning pipelines. – Taught supervised & unsupervised learning: regression, classification, clustering, association rules. – Built complete hands-on workflows using pandas, scikit-learn, regex, and Great Expectations. – Led interactive workshops (Slido quizzes), live coding demos, and collaborative corrections. – Supervised a final RH project (classification of contract types or employee segmentation).

– Designed and deployed secure AI agent systems to assist internal IT teams in their daily operations. – Developed AI agents (collaborative multimodal agents capable of performing tasks autonomously) based on LLMs, integrated into Microsoft Teams. These agents help IT support respond quickly to frequent requests (VPN configuration, password reset, tool access). – MCP Server to connect agents with internal tools – Conducted vulnerability testing (prompt injection) to strengthen the security and reliability of AI-generated responses. – Evaluated responses in collaboration with business teams while ensuring scalability and compliance requirements were met. – Collaborated with IT governance teams to define internal usage policies and implement performance monitoring tools. – Deployed solutions using Docker and FastAPI.

– Designed and implemented a large-scale Retrieval-Augmented Generation (RAG) system across several thousand documents, orchestrating multiple complementary AI agents to enhance information retrieval. – Deployed the solution within a general-purpose chatbot, enabling regional bank advisors to quickly access accurate information and improve customer support efficiency. – Built with LlamaIndex, Streamlit, FastAPI, Python, and open-source LLMs, ensuring scalability and adaptability to enterprise needs.

Machine Learning Engineer – MAIF (Niort) – Built and industrialized a scoring system to explain tariff evolution on home insurance contracts for policyholders. AI Engineer - Generative AI POCs to Industrialization – DIOR (Luxury): Retrained image generation models on visuals of the Dior J'adore perfume. From text prompts, generated new brand-consistent images in diverse visual contexts. – Micropole (Consulting): Developed a private ChatGPT for HR contracts, enabling employees to query past contracts and verify clauses before responding to RFPs. Data Scientist – LESIEUR (Retail / FMCG) – Developed a sales volume prediction model per client for Lesieur's product portfolio, improving stock management and optimizing client-specific sales. AUDI (Automotive) – Machine Learning Engineer: Deployed (AWS) a predictive tool for electric vehicle sales, enabling Audi teams to anticipate factory orders ahead of client purchases and reduce customer delivery delays. – Data Scientist: Improved customer experience on Audi.fr through a client segmentation project. – Led scoping workshop to define objectives, scope, and available data. – Built segmentation of customers configuring electric vehicles on the website to support personalized experience strategies.

Data Scientist – ENI (Energy) Customer Churn Prediction – Built predictive models to identify customers at risk of churn, enabling targeted marketing campaigns that reduced costs and increased profitability. – Merged multiple data tables and engineered new explanatory variables. – Conducted statistical analysis to identify peak churn periods and the main drivers of customer attrition. – Created churn prediction targets based on identified periods and applied SMOTE to rebalance the dataset. – Designed and developed result visualizations in Power BI and delivered executive reports with actionable insights. Data Scientist – GRT-Gaz (Energy Infrastructure) Predictive Maintenance – Modeled equipment failure rates to optimize preventive maintenance, supporting cost savings and operational reliability. – Delivered dashboards and storytelling reports for business stakeholders. – Applied algorithms including XGBoost, logistic regression, random forest, PCA, and MCA. – Tools: Python, R, GitLab, Power BI, PowerPoint.

– Connected, retrieved, and prepared large datasets for analysis. – Programmed in R and Python using AWS notebooks (Athena, Glue). – Selected and evaluated predictive models (XGBoost) to measure accuracy and optimize forecasting performance. – Built interactive dashboards in Power BI to present results to business stakeholders. – Authored detailed reports and participated in Data Science team meetings at DGII, as well as client workshops. – Tools & Technologies: Python (ML libraries), R, AWS (Jupyter Notebook, Athena, Glue), GitLab, SQL, Power BI, PowerPoint.

– Tutored high school students (junior and senior year) in various subjects. – Carried out this activity after classes, driven by a long-standing passion for sharing knowledge since middle school. – Found fulfillment in helping students succeed academically and grow in confidence.