Senior Research Scientist (AI)

TL;DR

Senior Research Scientist (AI/Multimodal): Leading the development of vision and multimodal models for document and media translation with an accent on post-training, reinforcement learning, and knowledge distillation. Focus on building evaluator models, optimizing model delivery pipelines, and integrating complex ML systems into production.

Location: Hybrid (London, UK)

Company

DeepL is a global AI product and research company specializing in secure, intelligent language AI solutions for translation and writing.

What you will do

• Drive the development of vision and multimodal models for document, image, and media translation.
• Lead research and development on post-training, including supervised fine-tuning, knowledge distillation, and RL (RLHF/RLAIF, PPO/GSPO).
• Build evaluator models for document and design quality and mitigate reward hacking and quality-estimation failure modes.
• Own the full model delivery lifecycle from prototyping and ablations to production deployment in real-time systems.
• Establish strong practices for evaluation, reproducibility, and continuous model improvement in production.

Requirements

• Proven experience developing multimodal models, VLM, and/or vision models.
• Deep expertise in model post-training, knowledge distillation, and reinforcement learning.
• Strong data-centric instincts for building synthetic-data and preference-data pipelines.
• Strong coding and experimentation skills in Python using PyTorch, JAX, or TensorFlow.
• Ability to lead complex research efforts and align research goals with product and engineering priorities.
• Must be based in or able to work from the London office in a hybrid setup.

Nice to have

• Experience with machine translation, multilingual NLP, or efficient long-context modeling.
• Knowledge of diffusion models, multi-objective optimization, or unified multimodal generation.
• Experience designing evaluation signals using Model-as-judge or human-in-the-loop evaluation.
• Publications at top-tier AI venues.

Culture & Benefits

• Highly collaborative environment working across research, engineering, and product teams.
• High-impact role with the scope to drive research breakthroughs directly into production.
• Company culture rooted in trust, curiosity, care, and a commitment to diversity.
• Opportunity to shape the future of AI within a purpose-driven, fast-moving organization.