Research Scientist at Atlys in Delhi HQ

Job description

About Atlys Our mission is to enable every person on Earth to travel freely. Travel should feel exciting. Instead, too much of it is still defined by paperwork, uncertainty, and processes that were never designed for travellers. Atlys is changing that. We're building an AI-first platform that removes friction and uncertainty from travel. To date, we've processed visas for 1M+ travellers across 127 countries. But we're still early. The Role We're looking for an Research Scientist to push the boundaries of what AI systems can actually do. Not someone who simply integrates existing models or builds another wrapper around an LLM. We're looking for someone who can identify why today's AI systems fail, develop new methods to overcome those limitations, and turn research breakthroughs into real-world capabilities. You'll work on some of the hardest problems in applied AI: building autonomous agents that can plan, reason, execute complex tasks and recover from failures; improving model intelligence through reinforcement learning and post-training; and developing systems that can make reliable decisions in unpredictable environments. We're looking for someone who doesn't just ask how to use the best available model, but how to make AI fundamentally better at solving problems it struggles with today. The Job • Push the boundaries of agentic AI. Research and develop new approaches that enable AI agents to independently plan, reason, use tools, execute complex tasks, adapt to unfamiliar environments and recover from failures. • Advance reinforcement learning and model training. Experiment with RL, fine-tuning, post-training techniques and model architectures to improve reasoning, decision-making and autonomous behaviour beyond what existing models can achieve. • Develop new methods for reasoning and multimodal intelligence. Investigate why models struggle with complex, multi-step problems and unstructured inputs, and design new algorithms and techniques to improve their capabilities. • Design rigorous experiments and AI evaluations. Build benchmarks, evaluation frameworks and experimental methodologies to measure model intelligence, reliability, generalisation and failure modes. Use findings to drive meaningful improvements. • Take original research from first principles to production. Identify unsolved problems, formulate hypotheses, build and test new approaches, and collaborate with Research Engineers to translate breakthroughs into scalable AI systems with measurable real-world impact. The Ideal Candidate You'll likely thrive in this role if you: • Have a PhD in Machine Learning, Artificial Intelligence, Computer Science or a related field, or equivalent experience conducting original AI research. • Have a strong track record of original research, demonstrated through publications at venues such as NeurIPS, ICML, ICLR, ACL or EMNLP, impactful open-source contributions, or meaningful advances in production AI systems. • Have deep expertise in one or more areas such as agentic AI, reinforcement learning, LLM reasoning, post-training, multimodal learning, model architectures or AI evaluations. • Are highly proficient in Python and PyTorch, with hands-on experience designing experiments, training and fine-tuning models, building evaluation frameworks and independently implementing research ideas. • Are intellectually curious, comfortable challenging existing approaches and excited by unsolved problems. You combine scientific rigour with strong engineering instincts and care about taking original research from experimentation to real-world impact. We hire exceptional people for their craft, not travel enthusiasts. We solve hard problems that happen to touch global travel and government infrastructure. What you can expect Atlys is built by a small team with outsized ambition, high ownership, and an obsession with doing great work. Here’s what you can expect: • Competitive compensation and equity so you participate in the value you help crea