Innodata is looking for exceptional minds thinking deeply about innovations in AI training and evaluation. Innodata’s frontier practices are building the teams, tools, data, and methods that develop, support, evaluate, train, and stress-test the world's most advanced AI models and agents. Innodata partners with leading frontier AI labs to push the boundaries of what AI systems can reliably do.

As an AI Researcher on our team, you will conduct cutting-edge research in reinforcement learning, agentic systems, and large language models. You will design methodologies that reveal genuine AI capabilities-not just surface-level pattern matching-and translate theoretical insights into scalable, production-grade systems that operate at scale on the frontier. You will wear multiple hats as a thought leader, educator, researcher, and developer on a small team at the cutting edge of the field within a growing industry leader.

What You'll Do

• Conduct novel research in AI training, evaluation, reinforcement learning, and multi-agent systems
• Design and implement evaluation frameworks for testing and improving reasoning, agentic behavior, and complex instruction-following in frontier models
• Prototype approaches, algorithims, and tools for verifiable rewards, RL, constraint-based testing, and interpretable evaluation metrics
• Analyze performance and scalability challenges in large-scale distributed evaluation and training pipelines
• Collaborate with cross-functional teams and external partners to bring research breakthroughs into production
• Contribute to Innodata's thought leadership in frontier AI training and evaluation methodologies

What You'll Bring

• MSc or PhD in AI, Machine Learning, Computer Science, Mathematics, or a related quantitative field
• Strong background in reinforcement learning, especially RLHF, verifiable rewards, or interpretable reward modeling, policy optimization, etc.
• Deep understanding of machine learning fundamentals-optimization, probability, statistics, and linear algebra
• Hands-on experience designing, implementing, and evaluating ML models or algorithms
• Ability to translate theoretical insights into scalable, production-grade systems
• Excellent analytical and problem-solving abilities with a creative and rigorous research mindset
• Excellent English communication skills

Bonus Points

• History of high-performance software development (Python, C++, or equivalent)
• Familiarity with agentic AI systems, multi-agent orchestration, or tool-use evaluation
• Experience scaling large models
• Familiarity with distributed training and inference frameworks (PyTorch, CUDA, Triton)
• Strong publication record or demonstrable research contributions at top venues (NeurIPS, ICML, ICLR, ACL)
• Domain expertise in STEM fields

Why Innodata

Innodata sits at the critical intersection of frontier AI development and rigorous evaluation. Our work directly shapes how the world's most capable AI systems are tested, refined, and deployed safely. You will work on problems that matter—designing the evaluation frameworks that determine whether AI systems can be trusted with complex, real-world tasks.

Our team includes researchers and engineers from academia and industry with diverse backgrounds. We value strong fundamentals, hands-on engineering, intellectual curiosity, and collaboration. This is not a narrow role—you will work across multiple domains, take real ownership, interact with customers, and be supported by teammates who care deeply about the work.

How to Apply

We believe strong teams are built from people with different experiences and perspectives. We encourage candidates from underrepresented groups in tech to apply, even if you don't meet every item listed above. Please provide your CV, transcripts (if relevant), relevant research (including dissertation or thesis), and a cover letter describing your interest in this role and why you think you would be a good fit (please don’t use an LLM for this portion). Email that to [email protected]