Luca Serra
Senior Machine Learning Engineer
Salaire / Taux journalier
Marseille, France
Mobilité : Paris, Lyon
Freelance
Peut venir sur site
Expertises
Langues
À propos
Introduction
Senior ML Engineer with 4+ years building production models for Fortune 500 companies including Decathlon, BNP Paribas and Sephora. Specialized in rapid deployment - delivered forecasting models in 3 weeks and object detection systems for industrial defect detection. Expert in MLOps, dynamic pricing, and AI chatbots with proven results at scale. More info on my website: https://luca-serra.com/
Expérience de travail
Machine Learning Engineer
Decathlon DigitalFreelance
mai 2024 - déc. 2024
8 mois
Paris, France
As the first MLE Engineer in the Pricing insights team, part of BCP (Business Capabilities Platform), my first goal was to take over an existing dynamic pricing project and deploy it in Decathlon Digital infrastucture (stack: AWS, Databricks, Airflow) on more than 10 countries. My impact on this project: • Deployed dynamic pricing script running every day on recommended infrastructure (Databricks, AWS MWAA, AWS ECR, AWS S3, AWS CA) • Drop of script failure rate by 30pts (now close to 0) • Enhanced monitoring and failure handling (alerting, retry...) • Implementing code best practices (Domain-Driven Design, CI/CD, unit tests...) and proposing state-of-the-art tools for code management and quality (e.g. uv, ruff) Also prepared the industrialization of a second dynamic pricing project, leveraging product similarity (embeddings, LLM) and collaborated with two data scientists.
Artefact
mai 2021 - mai 2024
3 ans 1 mois
Paris, France
Senior Data Scientist
janv. 2023 - mai 2024
1 an 5 mois
Paris, France
Missions: • @BNP Paribas: Built an online bank assistant chatbot using RAG (Retrieval Augmented Generation). • @Khimod: Implemented in a few weeks an object detection model to spot defects on metal surfaces of heat exchangers. Contributed to the repository Ultralytics. • @Sephora: Developed a forecasting model in just 3 weeks to help Sephora to estimate the sales in 5 countries 1 year ahead. • @Adeo/@Leroy Merlin: Built in a very short timeframe an assortment recommender web application used by 30+ department managers ; with a monitoring dashboard. Teaching Data science courses to several engineering schools: • MLOps at X-HEC • Viral marketing at CentraleSupélec • Data Science for business at École des Ponts ParisTech • Forecasting at Arts et Métiers Internal work: Developed several starter kits and accelerators for Artefact's internal developer portal.
Data Scientist
mai 2021 - janv. 2023
1 an 9 mois
Paris, France
Missions: • @Richemont: Industrialized and automatized two forecasting pipelines using GCP services (Vertex pipelines, Cloud scheduler, Cloud functions etc...) • @Fnac-Darty: Developed a counterfactual forecasting model to estimate past promotions uplift (linked article: https://t.ly/_mycU) • @Nestle: Data Scientist contributor on a POC • @Fortenova: Data Scientist lead on a POC Internal: • Active contributor to Artefact forecasting expertise. • Prepared a training session on the use of boosting models for forecasting (CatBoost).
Backend developer
Voyagez VaccinéFreelance
août 2020 - mai 2021
10 mois
Paris, France
Backend development of a mobile app (+ a bit of frontend), which helps and guides people who travel to a high-risk foreign country. A very unique trip experience can be achieved thanks to user's health data. Deployment of the app on a secure health-certified server. 💻: Node.JS, Express, PostgreSQL, Sequelize, React.JS
Formation
CentraleSupélec
Ecole Centrale Paris
2017 - 2021
3 ans 9 mois
The Faculty of Engineering at Lund University
Industrial engineering
2019 - 2019
1 mois
Lycée du Parc
Math & Physics
2015 - 2017
2 ans 1 mois