Umberto Cifani Stroebel

– Strategic AI & Risk Leadership: Led development and deployment of ML pricing models (Tweedie distribution) and underwriting systems, delivering AI-driven decision engines and risk analytics. Translated complex business requirements into scalable AI solutions, optimizing efficiency. – Market Intelligence: Directed competitor and KPI analysis to evaluate product performance, leveraging market trends to inform decision-making. – Leadership: Managed 3 people + 4 external consultants. Drove data partnerships with Serasa Experian. – Achievement: Deployed deep learning models on GCP in Python, reducing loss ratio by 1% (€6M) and significantly enhancing product profitability.

– Leadership: Managed a 15-person cross-functional team (Data Science, Data Engineering, and MLOps). – Product Strategy: Directed the end-to-end development and commercial launch of a new AI product from concept to market. – CRM Analytics Strategy: Led Marketing campaigns / KPI's using SQL and Power BI, driving advanced segmentation strategies that improved campaign ROI by 15%. – Achievement: Designed and delivered AI-powered customer segmentation software (XGBoost, Logistic Regression, Random Forest in Python on AWS Cloud), successfully launched commercially and adopted by multiple clients. – Integrated LLMs (ChatGPT) to automate contact center call auditing, reducing manual QA effort by 30% while improving quality consistency.

– Delivery Data Science Model (Machine Learning): Logistic Regression vs Artificial intelligence (AI) in Credit and Fraud score.

– Strategy: Led end-to-end Fraud Detection & AML ML strategy (SAS, Python) balancing fraud prevention, transaction growth, and customer experience. Delivered real-time alerting and executive KPIs using Tableau. – Leadership: Built cross-functional team from scratch 4 analysts (Data Scientists, Fraud Analysts). – Partnership: Defined and aligned fraud policy with Finance, Risk, and VP; ensured 100% Central Bank compliance. – Achievement: Deployed fraud detection infrastructure with real-time ML models and alerts rules. – Drove 40% transaction growth and 10% fraud loss reduction YoY while optimizing detection accuracy.

– Strategy: Developed annual Credit & Fraud scoring models (PD) using: SQL, SAS and Logistic Regression for B2B (Corporate, SME) and Retail; optimized collection policies and detection alerts/models/rules across Risk, Fraud and Operations. – Expertise: Technical leader for new financial product evaluation and profitability analysis; led cross-functional analytics projects spanning CRM and financial data. – Partnership: Strategic partner to Risk and Analytics leadership across Modelling and Credit Policy departments, bridging issuing and acquiring operations for integrated risk management. – Achievement: Delivered Credit & Fraud Score models (PD), increasing market penetration of new financial products while minimizing losses. – Uncovered multi-million € internal fraud scheme via analytics, triggering complete vendor outsourcing policy revision. – Also built an operational BSC, improving case analysis productivity by 40%.

– ML Models: Developed credit/behavior scoring models (SAS, SPSS) B2B/ B2C portfolio. Calculated limits and optimized credit policies. – Predictive Analytics: Forecasted financial outcomes and marketing plan impacts for long-term planning. – Team Leadership: Led a team of 2 credit analysts. – Achievement: Delivered first Credit & Behavior Score model: increasing approval rates by 15% while maintaining risk control (churn).

– Responsible for FCA CRM model; – Responsible for choosing the target audience for advertising campaigns; – Studying survey results and performance analysis related to customer profiles; – Leading a team of statistics analysts.

– Data Strategy: Developed clustering-based segmentation (k-means) and forecasting tools to drive marketing ROI. – Optimization: Built customer targeting and campaign analytics to maximize profitability. – Achievement: Increased sales by 15% through advanced customer segmentation.

Internship at Atos in Turin. Studied statistical models, Data Mining and Machine Learning (ML).