Data Driven Polling- Led a team in optimizing cache quality for shopping products with billions of transactions per day, resulting in improved platform quality and new features discovery.- Managed all aspects of the project, including data management, feature engineering, offline model evaluation, model tuning, and performance monitoring.- Migrated the data pipeline and model training from on-premise servers to the cloud, utilizing Azure, Kafka, Spark, Databricks, and MLflow.- Conducted deep dive sessions with colleagues to help them better understand the models’ decisions- Advocated for functional experts not familiar with machine learning and data to get access to raw data, resulting in new patterns being discovered and the platform quality improved.

Data Science Community NORAM- Started and animated the Data Science community NORAM (~100 members)- Organized a series of awareness sessions to educate members on the latest developments in data science, including study groups and coding dojos focused on key concepts and tools.- Provided feedback and support to newcomers and individuals starting with machine learning, fostering a supportive and collaborative environment for all members.Meta-connect rule engine- Implemented machine learning-based method to generate business rules for the meta-connect engine- Data collected and processed in Python, machine learning algorithm written in Spark- Rules are published for ease of access via a R-Shiny based web application

Clicks analysis for US Online Travel agency- Developed an innovative visualization and analysis tool to improve Master Pricer quality, resulting in improved accuracy.- Managed an A/B testing campaign to assess the impact of Instant Search results on conversion, optimizing product performance.Flex Pricer Proof of concept for a major US Airline- Led a successful task force to improve the response time for Flex Pricer Premium, producing clear and actionable weekly status reports for management.Affinity Shopper for 3 major airlines- Developed ad-hoc tools in Python to optimize the quality of the Affinity Shopper product, applying machine learning and data analysis techniques to support investigations.- Communicated and collaborated with customers, product and development teams, producing visualizations to share results.

- Designed and implemented from scratch the technical and functional monitoring of the SSP Test Systems (>200 servers, >200 internal users, customer facing systems, heterogeneous environments)Done with perl and Nagios- Hardware procurement management: from specification to validation and delivery to customers.- Supported all the SSP Tests environments. Problem and Incident management- Building, qualifying and planning of Pricing & Shopping Platform releases for production.- Demonstrated high level of proactivity, efficiency and autonomy

Main projects: Automatic Ticket Changer, Award pricer, Automatic Ticket Refund, Netting,mandates necessary for airlines.Coordination with different departments (marketing, pricing, technical services and ticketing).Procedure implementation, definition.Integration, optimization, regressions of releases for production loads.Problem fixing and change requests implementation according to plan and budget.Code and performance testing, validation.