Lyn Boussengui

• Developed and implemented explainable (glassbox) and efficient machine learning models for credit risk assessment for the largest credit insurance group • Conducted comprehensive data analysis and visualization as a precursor to model development • Engineered industrial-grade, optimized code to build a robust modeling platform • Deployed AI solutions using API and container technologies • Collaborated with data and infrastructure teams to establish an end-to-end ML pipeline and CI/CD processes • Effectively communicated and explained complex data science concepts to users and stakeholders • Operated in an agile environment, ensuring rapid iteration and continuous improvement • Contributed to enhancing the company's credit risk assessment capabilities through advanced analytics and machine learning techniques

• Led corrective, operational, and evolutionary maintenance of data science and AI/ML pipelines for the world's largest cosmetics group • Defined and implemented data acquisition and analysis strategies to support R&D initiatives • Developed an innovative optimization algorithm for adjusting chemical formulas based on machine learning model prediction • Applied and enforced internal guidelines for algorithm industrialization, including best practices for code development and data science processes • Collaborated cross-functionally with data teams, front-end and back-end developers, and UI/UX designers to establish robust end-to-end pipelines • Effectively communicated complex data science concepts to product owners and users, bridging the gap between technical implementation and business needs • Operated in an agile environment, ensuring rapid iteration and alignment with evolving R&D requirements • Contributed to advancing L'Oréal's R&D capabilities through the application of cutting-edge data science and machine learning techniques

Luxury and Cosmetics Industry: • Led efficient exploration of new formulation fields using Design of Experiments (DOE) and advanced machine learning models. • Spearheaded the automation of raw data acquisition, implementing robust data preprocessing pipelines • Developed machine learning models to predict colorimetric indicators based on chemical formula composition • Participated in a proof of concept applying Bayesian methodology to incorporate business knowledge into AI/ML models • Deployed AI/ML solutions using FastAPI and containerization technologies with Docker Athletic Shoes and Sports Equipment: • Established a comprehensive registration strategy for production-ready models within the Reusable team • Defined critical metadata parameters for model documentation, including data, code, parameters, hyperparameters, and ownership • Created a proof of concept for a model registry based on real-world use cases • Developed data science and code development guidelines documentation Additional Projects: • Built a regression model to analyze product price elasticity for the European leader in tire manufacturing • Remodeled network planning for notary establishment across French territories • Enhanced web API functionality using Django framework Key Achievements: • Successfully implemented end-to-end ML pipelines from complex data sources across multiple industries • Effectively communicated technical concepts to product owners and stakeholders • Gained extensive experience in agile environments and cross-functional collaboration

• Developed a machine learning model for anomaly detection among health professionals, focusing on opticians, using extensive health care data for France's largest insurer • Conducted deep statistical analysis on health data to understand and characterize fraud phenomena • Engineered relevant features through comprehensive descriptive statistical analysis • Evaluated and implemented advanced anomaly detection algorithms, including One-class SVM and Isolation Forest • Collaborated with cross-functional teams to integrate the model into existing fraud detection processes • Delivered a high-impact solution that generated €700,000 in profit during its first year of production • Gained hands-on experience in applying machine learning to real-world financial and healthcare challenges • Contributed to enhancing AXA's fraud detection capabilities, directly impacting the company's bottom line

• Contributed to developing a comprehensive customer churn prediction model (binary classification modeling) • Designed and implemented a robust data pipeline to extract and integrate information from diverse sources, including: - Customer account transactions - Credit card usage patterns - Online banking activity - Client communication history • Performed advanced feature selection using LASSO and Stepwise regression with bootstrap resampling in R • Developed and compared binary classification models using Logistic Regression and Random Forest algorithms • Collaborated with the wealth management team to align the model with business objectives and client retention strategies • Gained hands-on experience in applying data science techniques to real-world financial services challenges

• Conducted short-term forecasting of sectoral natural gas consumption in the Netherlands for ENGIE, the third-largest global group in the non-oil energy sector • Developed a two-week ahead prediction model to optimize energy distribution and resource allocation • Performed comprehensive descriptive statistics and statistical tests to analyze time series characteristics and determine optimal combinations of climate variables • Implemented and compared the performance of classical statistical models for time series against machine learning approaches: - Classical models: ARIMAX, SARIMAX - Machine learning models: Support Vector Regression (SVR), Extra-Trees • Collaborated with energy experts to incorporate domain knowledge into the forecasting models • Contributed to ENGIE's efforts in improving energy demand prediction and resource management

• Contributed to enhancing calculation tools for conducting annual statement information controls, improving accuracy and efficiency of financial reporting processes • Developed VBA programs to automate key processes, significantly reducing manual workload and potential for human error • Actively participated in testing and validation of tool enhancements, ensuring robustness and reliability of developed solutions • Collaborated with cross-functional teams to identify areas for improvement in existing workflows and implement data-driven solutions • Gained hands-on experience in applying programming skills to solve real-world financial and operational challenges

• Conducted comprehensive data analysis on a nationwide survey for the National Center for Scientific Research's laboratory on social issues (IRIS) • Analyzed public opinions and behaviors on critical societal themes including taxation, policing, education, and justice • Implemented advanced statistical techniques to extract meaningful insights: - Performed Multiple Correspondence Analysis (MCA) to identify patterns in categorical data - Utilized Hierarchical Agglomerative Clustering (HAC) to group individuals based on similar behaviors and opinions - Employed Logistic Regression to analyze odds ratios, quantifying the strength of association between variables • Synthesized complex findings into actionable insights, contributing to a deeper understanding of French public opinion on key social issues • Collaborated with senior researchers to interpret results and prepare reports for stakeholders

• Collaborated with the Internal Audit and Control Manager to evaluate health training costs, contributing to resource allocation optimization in the healthcare education sector • Conducted comprehensive statistical analysis on survey data using advanced multivariate techniques: - Performed Multiple Correspondence Analysis (MCA) to explore relationships between categorical variables in healthcare training - Implemented Multiple Factor Analysis (MFA) to analyze complex datasets with multiple groups of variables - Utilized Hierarchical Clustering on Principal Components (HCPC) based on MCA results to identify distinct groups and patterns in the data • Synthesized complex statistical findings into actionable insights for university administration and healthcare education stakeholders • Gained hands-on experience in applying advanced statistical methods to real-world challenges in healthcare education management

• Contributed to the preparation of periodic financial reports, ensuring accuracy and compliance with banking regulations and international financial reporting standards • Maintained and updated critical accounting records, supporting the bank's financial integrity and transparency • Provided daily administrative support to senior management, enhancing operational efficiency and effectiveness • Conducted customer follow-ups, improving client relations and supporting the bank's customer retention efforts • Participated in the development and implementation of the ARIZ AFD (Agence Française de Développement) warranty eligibility grid, contributing to risk assessment and management processes • Gained comprehensive exposure to banking operations, financial analysis, and risk management in an international banking environment

• Conducted comprehensive descriptive statistical analysis on the BAAC (Bulletin d'Analyse des Accidents Corporels) database, contributing to national road safety research • Developed and implemented key statistical indicators for CEREMA's annual publication on road safety, enhancing public understanding of traffic accident trends • Performed logistic regression analysis to investigate the impact of driver fatigue on traffic accidents in 2012, providing valuable insights for road safety policy-making • Collaborated with interdisciplinary teams of researchers and policy makers to translate complex statistical findings into actionable road safety recommendations • Contributed to CEREMA's mission of improving public safety through data-driven insights and evidence-based policy recommendations

• Contributed to the design and implementation of a comprehensive research questionnaire, focusing on marine and coastal resource management • Conducted in-depth statistical analysis of survey data, providing valuable insights into resource usage patterns and stakeholder perspectives • Estimated economic gains from tourist attractions in coastal areas, contributing to the understanding of the economic impact of marine and coastal tourism • Collaborated with senior researchers to interpret results and prepare reports for stakeholders in marine resource management and coastal tourism • Participated in interdisciplinary research projects, bridging the gap between environmental science, economics, and public policy