AI Research Scientist (Paris)

Lexsi Labs; Paris; Sur place; EUR 85 000 - 130 000; Plein temps

Résumé du poste

Lexsi Labs is at the frontier of building aligned, interpretable and safe AI, with a focus on scalable ML/DL methods. You will tackle advanced AI safety, explainability, and alignment problems in large-scale settings, while collaborating with MLEs and SDEs to ensure reliability and smooth deployment.

The role emphasizes publishing peer‑reviewed papers, presenting at industry venues, and contributing to open research culture in a flat, fast‑paced environment in Paris.

Qualifications

• PhD or MSc in CS or Mathematics with strong ML foundations.
• Peer‑reviewed papers or significant open‑source contributions.
• Hands‑on DL frameworks experience (TF, PyTorch).
• Experience applying ML to diverse data types (text, tabular, image).
• Cloud/on‑prem deployment experience for ML models.
• Solid grounding in ML explainability methods.
• 2+ years in DL/ML and strong research track record.
• Experience with Transformer, GANs, and other DL techniques.

Responsabilités

• Advance AI alignment, safety, and explainability research.
• Publish papers and present at conferences.
• Collaborate with ML engineers to deploy ML features.
• Maintain technical and product documentation.
• Experiment with diverse data types to improve models.

Connaissances

ML concepts

Deep learning

Reinforcement learning

MLOps basics

Explainability (LRP/LIME)

Transformer models

GANs

Python

Cloud deployment

Publications

Formation

Master's or PhD in CS/Math

Outils

TensorFlow

PyTorch

Description du poste

Lexsi Labs is one of the leading frontier labs focusing on building aligned, interpretable and safe Superintelligence. Most of the work involves on creating new methodologies for efficient alignment, interpretability lead-strategies and tabular foundational model research. Our mission is to create AI tools that empower researchers, engineers, and organizations to unlock AI's full potential while maintaining transparency and safety.

Our team thrives on a shared passion for cutting-edge innovation, collaboration, and a relentless drive for excellence. At Lexsi.ai, everyone contributes hands‑on to our mission in a flat organizational structure that values curiosity, initiative, and exceptional performance.

As a research scientist at Lexsi.ai, you will be uniquely positioned in our team to work on very large-scale industry problems and push forward the frontiers of AI technologies. You will become a part of the unique atmosphere where startup culture meets research innovation, with key outcomes of speed and reliability.

Responsibilities

• You’ll work on advanced problems related to AI explainability, AI safety, and AI alignment.
• You’ll have flexibility in picking up the specialization areas within ML/DL and problem types that address the above challenges.
• Create new techniques around ML Observability & Alignment.
• Collaborate with MLEs and SDE to roll out the features and manage their quality until they are fully stable.
• Create and maintain technical and product documentation.
• Publish papers in open forums like arxiv and present in industry forums like ICLR NeurIPS etc.

Qualifications

• Has a solid academic background in concepts of machine learning or deep learning or reinforcement learning.
• Master or Ph.D in key engineering topics like computer science or Mathematics is required.
• Should have published peer‑reviewed papers or contributed to open‑source tools.
• Hands‑on experience in working with deep learning frameworks like Tensorflow, Pytorch etc.
• Enjoys working on various DL problems that involve using different types of training data sets – textual, tabular, categorical, images etc.
• Comfortable deploying code in cloud environments/on‑premise environments.
• Good fundamentals in MLOps and productionising ML models.
• Prior experience on working on ML explainability methods – LRP, SHAPE, LIME, IG, CEM etc.
• 2+ years of hands‑on experience in Deep Learning or Machine Learning.
• Hands‑on experience in implementing techniques like Transformer models, GANs, Deep Learning, etc.