Design, develop & deploy ML/LLM models, data pipelines & agentic AI solutions for int/ext projects

You will join our DataScience team, which focuses on the research and development of AI solutions built on DL, NLP, Computer Vision,LLMs and agentic AI.

We work on cuttingedge AI projects, tackling complex challenges related to the processing of textual and visual data in order to meet the needs of a variety of business use cases.

The team drives projects from research to production, collaborating actively with a range of internal and external teams.

Working as part of the team, you will contribute to the following activities:

Analyze complex business and technical problems and propose innovative AI solutions.

Lead benchmarking and evaluation studies to assess models, algorithms, and approaches.

Design, implement, and evaluate machine learning and deep learning algorithms.

Process and leverage large volumes of unstructured data, including text and images.

Develop maintainable, tested, production-grade Python code.

Optimize DL and LLM models for scalability, efficiency, and performance.

Productionize algorithms and models as robust, scalable API services.

Containerize and deploy AI services to cloud environments.

Write and maintain technical and scientific specifications, documentation, and experiment reports.

Proficient in Python: Strong Python skills, with hands-on experience using Deep Learning and LLM frameworks and libraries such as PyTorch, TensorFlow, and Hugging Face.

Strong ML fundamentals: Solid understanding of neural network architectures, machine learning techniques, and methods for data analysis, processing, and transformation.

Business-oriented mindset: Interested in solving concrete business problems and applying AI expertise to a variety of real-world use cases.

Cloud & deployment: Knowledge of deploying AI solutions to the cloud, including Docker, Kubernetes, cloud services, and related infrastructure, is a strong plus.

Production-grade development: Able to write clean, maintainable, tested, and production-grade Python code.

Curiosity & ownership: Motivated to learn, organized, curious, open-minded, and proactive, with strong problem-solving and interpersonal skills.

Teamwork: Able to work effectively in a collaborative, multidisciplinary, and innovative environment.

Languages: Fluent in French and English, both written and spoken. Daily team communication is in English.

Technical Stack

Language & frameworks: Python 3.11, PyTorch

GPU acceleration: CUDA, NVIDIA GPU drivers

ML/LLM libraries: Hugging Face Transformers, torch.nn, ONNX, etc.

LLM inference & serving: vLLM, NVIDIA Triton Inference Server

Domains: NLP, Computer Vision, state-of-the-art (SOTA) LLMs

LLM applications: RAG, Graph RAG, Vector Databases

Experiment tracking: MLflow, Weights & Biases (W&B)

Containerization: Docker; Kubernetes is a plus

APIs: REST

Infrastructure & tooling: Linux, Git, CI/CD