Description

Ingénieur Full Stack depuis 7 ans, je travaillais dernièrement au New York Times où j'ai mené plusieurs développements autour de l'IA:

- Un outil de ciblage publicitaire utilisant l'IA pour un matching sémantique profond de campagnes avec des articles de presse.

- Développement d'une plateforme IA entreprise: embedding pipeline, bibliothèque de modèles IA exposés en endpoints, banque d'embeddings,

- Développement d'une plateforme de test et de training de modèles transformers à destination des Data Scientist.

Industry field of expertise

Languages

French
Native or bilingual
English
Native or bilingual
Spanish
Conversational

Workplace preferences

Can work on-site

Paris 1er Arrondissement (up to 50km)

The New York Times
Senior Data Engineer
December 2022 - Today (3 years and 6 months)
New York, NY, USA
Ads Targeting Team: Designed and implemented ML pipelines for contextual and behavioral targeting. Embedding search and storage, LLM, distributed systems.
Acast
Senior Software Engineer
September 2020 - December 2022 (2 years and 3 months)
United States
Payments: Implemented services to handle daily 10 000's of CB transactions through payment gateway services.
Webedia
Software Engineer
June 2018 - July 2020 (2 years and 1 month)
New York, NY, USA

• Set up dynamic branching to run in parallel hundreds of subDAGs generated from a batch of argument sets stored in PostgresQL tables and updated after completion.
• The operational tables are mapped through foreign key references to encyclopedic tables meant to store the imported data. Those tables are exposed through a G raphQL API which cleverly represents the movie universe graph .
• Architected and deployed the entire cluster on Google Cloud Platform using Kubernetes to orchestrate the Docker containerized modules of the system and allow automatic scaling controlled by the scheduler .
• Implemented Kubernetes Executor spinning up an ephemeral container for every task, replacing the Celery task listener which results in optimizing cluster resource consumption .
• Developed a tailored monitoring tool with Next.js to fetch a high amount of data inside the cluster before rendering it in a UI with dynamic data visualizations.