Vincent Labarre Email & Phone Number

Villeneuve-Loubet, Provence-Alpes-Côte D'Azur, France

Valbonne, Provence-Alpes-Côte D'Azur, France

Continuation of the projects I was working on at Atos with more technical and supervisory roles.

Valbonne, Provence-Alpes-Côte D'Azur, France

Codex Dataplatform: • Development of a 100% cloud-native open-source data platform based on Kubernetes. • Use of CICD and GitOps principles, enabling IaC (Infrastructure as Code) • Adding/Updating/Exploring/Exploiting open-source components, Improving code, Adding features for deployment • Types of Components: ◦ Orchestration (Argoworkflow, Argoevents), ◦ Monitoring (Grafana, Loki, Prometheus), ◦ Data Visualization (Superset) and… Show more Codex Dataplatform: • Development of a 100% cloud-native open-source data platform based on Kubernetes. • Use of CICD and GitOps principles, enabling IaC (Infrastructure as Code) • Adding/Updating/Exploring/Exploiting open-source components, Improving code, Adding features for deployment • Types of Components: ◦ Orchestration (Argoworkflow, Argoevents), ◦ Monitoring (Grafana, Loki, Prometheus), ◦ Data Visualization (Superset) and Governance (Datahub), ◦ Processing (Kafka, Airbyte), ◦ Storage (Minio, Clickhouse), ◦ MLOps (Kubeflow, MlFlow, Feast)VDM (Ville De Marseille; City of Marseille):Development and maintenance of the M-Pulse-BDA application on the Atos Codex platform. This application displays on a map of the city of Marseille the various events (from internal sources of VDM services for the batch ingestion part and from Waze flows for the streaming ingestion part) taking place there, and assigns them a risk score based on a GDBT (GraDient-Boosted Trees) Machine Learning model.A public version can be consulted on the website (https://m-pulse.marseille.fr/): • taking into account changes to source data for ingestion, • adding functionality for accessing Data or displaying on the Front. • debugging application issues (ingestions, compute workflows, ...).Languages used: • Python with Spark for ingestions. • SQL Clickhouse for querying data. • Bash for deployment scripts. Show less

Sophia-Antipolis

• State of the art on Federated Learning (FL) and implementation at hand and with flower.• Development of a FL application called FedAL (Federated Atos Learning): • Containerization and Portability on Docker-Compose. • Render the application Cloud-Native with Kubernetes. • Integration to Atos Codex Data Platform.• Study on the functioning of the digital health ecosystem in France and the constraints in the use of medical data (anonymisation, pseudonymisation… Show more • State of the art on Federated Learning (FL) and implementation at hand and with flower.• Development of a FL application called FedAL (Federated Atos Learning): • Containerization and Portability on Docker-Compose. • Render the application Cloud-Native with Kubernetes. • Integration to Atos Codex Data Platform.• Study on the functioning of the digital health ecosystem in France and the constraints in the use of medical data (anonymisation, pseudonymisation, …)• Study on the human and technical governance necessary for training and inference with a centralized/decentralized monitoring Show less

• A 3-month internship at EURECOM with Huawei Munich Research Center in Munich. • Analysis of traffic data from cameras located in the Bantian station in Shenzhen, China• Study of the SUMO (Simulation of Urban MObility) traffic simulator: how it works, what it takes as input, what it can give as output. • Analysis in detail of the outputs and comparison of the consistency of the latter with the real value, by, in addition to traditional quantitative comparison methods, applying a… Show more • A 3-month internship at EURECOM with Huawei Munich Research Center in Munich. • Analysis of traffic data from cameras located in the Bantian station in Shenzhen, China• Study of the SUMO (Simulation of Urban MObility) traffic simulator: how it works, what it takes as input, what it can give as output. • Analysis in detail of the outputs and comparison of the consistency of the latter with the real value, by, in addition to traditional quantitative comparison methods, applying a machine learning method : the permutation test with MMD (Maximum Mean Discrepancy). Show less

Suresnes, Île-De-France, France

Subject: Research of factors (demographic, biological, medication...) that can influence hypothermia during a clinical stay.• Data analysis & visualization, • Univariate and multivariate analyses on a database containing numerous variables and sensitive medical observations using R.

Data Science And Engineering

Data Science And Engineering

Mathematics.Physics.Chemistry - Classes Préparatoires Aux Grandes Ecoles (Cpge)

PCSI (First year): First of promotion / PC (Second year): Top three

Mathematics.Phsyics.Chemistry - Baccalaureate Diploma, With High Honors (17,57/20)