Cyrille Delabre

- Led distinct teams of engineers and analysts, driving the overall data strategy. - Advocated for and promoted innovation through data science within the company. - Spearheaded the development of large-scale machine learning systems, including recommendation engines, fraud detection, and large language model (LLM) based systems. - Established research collaborations with public institutions such as ENS and INRIA, focusing on data and social science problems. Tech stack: Python • GCP (Big Query, Vertex, Cloud Run, K8s) • TensorFlow • Airflow / DBT • Clickhouse • Metabase

- Built a modern data stack infrastructure from the ground up and successfully deployed its key components. - Executed Sales and Operations Planning (S&OP) and forecasts, implemented anomaly detection algorithms, and established alerting systems for critical metrics. Tech stack: Python •Trino / Airflow • Metabase

- Promoted best practices in data science by developing machine learning tools for comprehensive data checks, efficient pipelines, and robust monitoring systems within the data org. - Developed machine learning models to attribute direct marketing efforts to key performance indicators, enhancing financial forecasting accuracy. - Implemented machine learning and API systems for anomaly and fraud detection, significantly reducing disruptions and minimizing losses from promotional costs. Tech stack: Python, Scala, Spark, Hive, Docker, Kubernetes, GCP, Hadoop, TensorFlow, Scikit-learn, Poetry, FastAPI, Tableau

• Built data pipelines for company’s reporting, analytics UDF functions & advanced Tableau dashboarding. • Built semi-real time ML fraud detection models through slack alerting; reduced fraud costs & time spent by 80%. • Increased company’s retention rate by 8% with mobile push automation & ML survival models. Tech stack: Python, Java, SQL • Hive, Spark • Tableau

• Built a web app to predict (with success) the outcome of soccer matches in UEFA Euro 2016 (data-scrapping, ML training & tournament generation).