ML Research Engineer

Overview

We are representing a well-funded AI safety organization in Paris that is building policy enforcement and optimization systems for large language models. Our client processes over 100 million API calls monthly and trains proprietary LLMs optimized for safety, speed, and cost. We are seeking multiple ML Research Engineers to join their core training team working on supervised fine-tuning, reinforcement learning from human feedback, direct preference optimization, reward modeling, and distributed GPU training.

Responsibilities

• Train and post-train large language models using modern alignment techniques including SFT, RLHF, and DPO
• Build and maintain large-scale data pipelines processing petabytes of text data
• Develop reward models and evaluation frameworks for model performance
• Implement distributed multi-GPU training infrastructure
• Build scalable representation pipelines including sampling strategies, preprocessing, embeddings at scale, and retrieval systems
• Apply LLMs for labeling, classification, weak supervision, and data enrichment with cost and quality controls
• Create self-serve analytics tools, datasets, and dashboards for team exploration
• Deliver actionable insights that influence product and operational decisions
• Collaborate with engineering teams to integrate model outputs into production workflows

What We're Looking For

• Strong proficiency in Python and SQL with production engineering experience
• Solid applied NLP and ML experience on real-world text datasets
• Hands-on experience with embeddings, clustering, topic modeling, semantic search, and classification
• Experience with distributed processing, large-scale data systems, and performance-cost optimization
• Knowledge of evaluation methodologies: offline/online metrics, human-in-the-loop labeling, inter-annotator agreement, drift monitoring
• Ability to build reliable data pipelines beyond notebook prototypes
• Understanding of ML failure modes and debugging techniques

Highly valued:

• Public builder footprint: open-source models, datasets, or frameworks on HuggingFace/GitHub
• Published benchmarks, research papers, or technical blog posts
• Experience training models at frontier labs or leading open-source model releases
• Knowledge of advanced RL methods beyond standard RLHF (online RL, GRPO-style methods, novel alignment approaches)
• Experience with moderation, safety, or classification models at scale
• Multilingual model training background
• Prior work with safety datasets, policy systems, or high-volume observability

How We Work With You

When you apply through Get A Job.ai, our talent team will review your profile and conduct an initial screening. If there's a strong match, we will coordinate your interview process with our client, which includes a take-home technical task, a technical interview with their Head of Applied Research, and a final conversation with leadership. We ask that you do not contact the client directly—all communication flows through our recruiting team to ensure a smooth, professional process.

Location and Benefits

This role is based in Paris with hybrid work arrangements. A relocation package is available for candidates moving to Paris. Our client offers comprehensive medical insurance for France-based team members, paid time off in line with local regulations, meaningful equity participation, all necessary hardware and tools, covered subscriptions for AI development tools, and team offsites twice yearly.

Pay

Compensation details will be discussed during the screening process based on experience level and qualifications.

Equal Opportunity: Get A Job.ai is committed to inclusive recruiting practices. We welcome applications from candidates of all backgrounds and work with our clients to ensure fair hiring processes.