Providing transparency over farming practices and farming carbon emissions using satellites and AI.

Architect, Head Data Scientist and Manager, I lead a team of 15 data scientists and engineers. With the right people, technologies and practices, I ensure that we deliver effective and maintainable code that makes a real business impact.Adapting to new opportunities, we leverage local and offshore hiring markets to bring the needed competencies to our team. With the right mix of engineers and scientists, we are able to deliver often through continuous integration. This has allowed us to adopt an agile setup, working in tribes, squads and chapters.As a data science architect, I led our team into the cloud—Google Cloud Platform. Exploiting native technologies in SaaS—BigQuery, Airflow, PubSub and AppEngine to name a few—and microservice architecture patterns, we have reduced development time, increased maintainability and improved performance several fold.On a day-to-day basis, our team finds and deploys data-driven solutions to a range of problems:• churn, up-sell and cross-sell scoring for 400k+ clients• acquisition models for 5M potential future clients• segmenting 30M+ active users with GPS and search data• advertising campaign automation and optimisation• optimisating search engine results ranking (PagesJaunes)Data technologies: Google Cloud Platform, Python, BigQuery, Airflow, Apache Beam (Dataflow), Spark (Dataproc), Kubernetes, Docker, AppEngine, Cloud Run, Kafka, PubSub, Tableau, PostrgeSQLData science technologies: MLflow, sklearn, SparkML, FFM, AI Platform

I led a team of data scientists to develop innovative digital marketing products and improve existing ones, taking responsibility for all stages from product conception and data collection to production. • Targeting users and measuring drive-to-store conversions using machine learning and over 9 billion GPS points / month • Automation of 100,000 RTB (real-time bidding) advertising campaigns • Segmenting 30M monthly users using search • Optimisation of click and drive-to-store conversion generation (+300%) in RTB marketsAwards: first place at "La Nuit des Rois 2020" for Local Impact in the section "Commerce connecté & Drive-to-store"Technologies: Python, Spark, Airflow, PostgreSQL, Tableau, Kafka, ElasticSearch, Mesos.

Solocal manages a portfolio of some of the most frequented web domains in the French market. These provide high quality information about users' preferences and inclinations.

Investments & Value ManagementShort-term contract

• Headed statistical analysis and portfolio modelling for NWQ's major investment offerings.• Structured derivative products.• Created a series of proprietary data visualisation and statistical tools fundamental to daily fund management.