Research Scientist

DeepL is a global communications platform powered by Language AI. Since 2017, we’ve been on a mission to break down language barriers. Our human-sounding translations and intelligent writing suggestions are designed with enterprise security in mind. Today, they enable over 100,000 businesses to transform communications, reach new markets, and improve productivity. Our goal is to become the global leader in Language AI, building products that drive better communication, foster connections, and make a real-life impact.

The Role

We are looking for passionate Research Scientists to join our core AI pillars. This unified role covers three of our most critical research areas. Depending on your expertise and interest, you will join one of the following teams:

• Language AI: Building the world's leading translation and text-improvement systems, taking responsibility for the entire model lifecycle from data to deployment.
• Foundation Model Task Adaptation (FMTA): Shaping how our models learn beyond pre-training. You will focus on RLHF, alignment, and post-training to enable new reasoning and controllability capabilities.
• Voice AI: Solving the "art of the possible" for real-time voice communication, including transcription, speech-to-speech translation, and low-latency audio generation.

Responsibilities

• Innovate & Build: Design and deploy state-of-the-art AI models — whether for translation, large-scale model alignment (RLHF/RLAIF), or multi-modal voice processing.
• Scale at Speed: Train neural networks at scale on DeepL's dedicated GPU clusters, pushing the boundaries of performance while maintaining low-latency production environments.
• End-to-End Ownership: Manage the entire lifecycle of research from theoretical modeling and prototyping to ablation studies and production deployment.
• Collaborate Globally: Work with ML Platform, HPC, and DevOps teams to integrate research into robust infrastructure serving millions of users.
• Advance the Field: Lead research initiatives that improve mathematical understanding of neural networks, ensuring reproducibility and high scientific standards.
• Operational Excellence: Participate in shared on-call rotations (specific to Voice/Production teams) to ensure reliability of global AI services.

Qualities we look for

• Technical Foundation: A solid mathematical background with a PhD, Masters, or equivalent industry experience in Computer Science, Mathematics, Physics, or related field.
• Engineering Proficiency: Deep practical experience in Python and at least one modern framework (PyTorch, JAX, or TensorFlow) evidenced through significant research projects or internships.
• Research Track Record: History of leading self-directed research projects that deliver tangible results including academic publications.
• Domain Expertise: Specialized experience in at least one of the following is beneficial:

• Large-scale LLM post-training and alignment (RLHF, RLAIF, RLVR)
• Neural Machine Translation (NMT) and text modeling
• Voice/Audio modalities (ASR, TTS, or Speech-to-Speech)

What we offer

• Diverse and internationally distributed team: part of a global community with more than 90 nationalities, including teams in the UK, Germany, the Netherlands, Poland, the US, and Japan.
• Open communication and regular feedback: clear, honest communication and collaborative culture that values empathy and growth.
• Hybrid work and flexible hours: office twice a week with flexibility to work remotely and adapt to time zones.
• Virtual Shares: ownership mindset with equity-like rewards linked to DeepL’s growth.
• Regular in-person team events and monthly hacking sessions: opportunities to collaborate and innovate with other teams.
• Annual leave: 30 days off (excluding public holidays) with access to mental health resources.
• Competitive benefits: location-aware benefits package that reflects a global team.
• Equal opportunity statement: We are an equal opportunity employer. You are welcome at DeepL for who you are—we appreciate authenticity here.