Greater Paris Metropolitan Region
Data & AI Consultant with 3+ years of experience designing and delivering AI and data solutions. Combines hands-on engineering expertise in Python, PySpark, APIs, cloud deployment and CI/CD with strong consulting skills in problem framing, stakeholder alignment and business value creation. Experienced in translating regulatory, risk and operational challenges into scalable AI/data architectures, with growing specialization in GenAI and agentic AI use cases from PoC to industrialization.
Advise business and technology teams on the design of AI, GenAI and agentic AI solutions addressing risk, regulatory monitoring, internal control and AML/CFT challenges. Partner with senior stakeholders across Data Office, Financial Security and Internal Control at Allianz France to frame business needs, define target solutions and translate requirements into scalable Data & AI architectures. Designed solution architectures combining data pipelines, scoring engines, RAG components, orchestration layers and workflow automation to support risk monitoring, analyst decision-making and operational efficiency. Supported business development efforts by structuring AI use cases, technical approaches, implementation roadmaps and value propositions for client proposals. Led internal GenAI enablement initiatives within Deloitte STeeR, delivering training sessions for 30 participants and knowledge-sharing presentations for audiences of up to 200 people on AI prompting and GenAI adoption.
Designed and deployed a cloud-native data validation solution on GCP, integrating Cloud Build, Cloud Run and CI/CD workflows to automate data quality controls. Implemented a data discovery, detection and classification infrastructure using Google Cloud DLP and Dataplex, supporting stronger data governance and sensitive data management. Built an NLP-based drift monitoring solution for Orange MEA chatbots, analyzing variations in user intents and n-gram distributions with spaCy and NLTK. Achieved 5th place among 30 teams in an electricity price forecasting hackathon, using time-series forecasting techniques such as ARIMA, Prophet and Darts.
Designed and implemented a price estimation solution based on machine learning regression models, leveraging commercial data from Salesforce to support pricing decision-making. Applied models such as XGBoost and scikit-learn regressors to estimate prices and identify key variables influencing prediction performance. Created interactive Plotly dashboards to visualize price estimates, model behavior and business-relevant pricing patterns. Assessed data quality issues affecting model performance and delivered recommendations to improve data completeness, consistency and predictive accuracy.
Designed and implemented a classification model to support card upgrade targeting on the B2C banking market, leveraging customer data and machine learning techniques. Processed and structured raw CSV data to create reliable datasets for feature engineering, model training and visualization. Applied XGBoost and scikit-learn classification models to identify customers with a higher propensity to benefit from a premium card offer. Created interactive Plotly visualizations to enhance customer knowledge and support business interpretation of model outputs. Used SHAP analysis to explain prediction drivers and improve the transparency of customer targeting recommendations.