Soukaina DIDI ALAOUI

Senior Data Scientist at La Banque Postale's Data Factory. Data Factory is a cross-functional department within the bank, positioned as a Data solutions center. Missions accomplished : • Identity Fraud Detection - Deep Learning : Project enabling the detection of fraudsters using the same picture on different fraudulent ID documents: vectorization of images & distance computation, development of a GUI, management of fraudster groups (Networkx). • ESG - Data Science/Engineer - design & development : Design of data solution allowing data collection from providers on Legal entities (PMO), data transformation and ingestion into the datalake. Development of a smart engine to reconcile this new data with the internal one. Thus allowing the calculation of ESG score for each Legal entity. • Product appetence score - Machine Learning - Industrialized : Implementing a machine learning model able to learn behavior of legal entities profils to improve product recommendation for marketing campaigns. • AML/CFT risk profile - Industrialized : Development of risk scoring for Legal Entities (companies, organizations,...) AML/CFT = Anti-Money Laundering/Combating the Financing of Terrorism • Business Layer design - Data Engineering - industrialized : Design and development of various business layers for the bank and its subsidiaries. • Textual matching (retail) - Data Analysis/Science - industrialized : Use-case 1 : Duplicates detection in databases (custom fuzzy matching). Use-case 2 : Enrich internal retail database with INSEE's open-source data. • FATCA EAI - Data Analysis - industrialized: Implementation of SAS-SQL scripts for monthly/annual declarations for the tax department. • Management : - Supervision of Data team (interns and juniors) - Ensure the comprehension of business needs - Ensure a good modeling of data products - Supervision of projects deployment & incidents management Role Progression From intern to Junior to Senior data scientist

Six-month internship in the Data Valorization and Enhancement department at La Banque Postale. This role required a mix of autonomy and collaboration, leveraging tools and languages such as Python, PyTorch, Dataiku (Big Data platform), SAS, and SQL. Main accomplished missions 🌟 Conducted extraction, exploration, and analysis of large datasets from diverse sources to address a variety of business requirements. 🌟 Developed and applied unsupervised models and techniques for fraud detection, including clustering and isolation forest methods. 🌟 Developed and implemented an initial Proof of Concept (POC) leveraging Deep Learning models to detect identity fraud by analyzing images of ID documents.