Research Scientist

This role is for someone with deep research expertise who is energised by problems that don't have a paper to copy yet, and who measures success by real-world impact, not just benchmark scores.

The Job

Set and drive a research agenda on open problems in agentic AI, including how LLM-powered agents plan, use tools, take actions and recover from errors over long, multi-step tasks.

Develop new methods for multimodal understanding and reasoning, so models can reliably extract, verify and reason over noisy, unstructured real-world inputs across formats and languages.

Design and run post-training research, including fine-tuning, preference optimisation and reinforcement learning, to push model capabilities beyond what general-purpose frontier models offer.

Research calibration, uncertainty estimation and self-verification so AI systems know when to act autonomously and when to defer to a human.

Build evaluations, benchmarks and synthetic data methods that give an honest picture of model quality, robustness and failure modes, especially on the long tail.

Collaborate closely with research engineers to take ideas from prototype to production, and share findings through internal write-ups and, where appropriate, publications.

The Ideal Candidate

You'll likely thrive in this role if you:

Have a PhD or MS in Computer Science, Machine Learning or a related field, or equivalent research experience, plus 3+ years of applied or academic research in ML, NLP or computer vision.

Have a track record of original research, such as publications at top venues (NeurIPS, ICML, ICLR, ACL, EMNLP, CVPR), influential open-source work, or research that clearly moved production metrics.

Have deep hands-on experience with LLMs beyond prompting, including fine-tuning, RLHF/DPO or other post-training methods, evaluation design, and a strong grasp of why models succeed or fail.

Are strong in Python and PyTorch, and can run experiments yourself, from data curation and training runs to careful analysis.

Bring rigour with a bias to impact: you ask sharp questions, design clean experiments, avoid fooling yourself, and care about whether the result changes what we build.

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 create.

Comprehensive health insurance covering you and your family, plus annual preventive health check-up on us.

A thoughtful leave policy with paid time off and company holidays, because great work also requires real time off.

Lunch, high tea and dinner at the office, so the basics are taken care of while we build together.

Learning & development credits to invest in courses, books, conferences and tools that help you get better at your craft.

Relocation support to make moving closer to the team as seamless as possible.

Great tools to do your best work, including high-spec equipment, software and compute for your research.

An office designed for more than just work, with pickleball, table tennis, PlayStation, an arcade game room, an F1 simulator and even a full flight simulator.

Our values shape how we make decisions, work together and build for travelers. The bar is high, the pace is fast, and the way we work is deliberate. Read them before you apply. If this environment energizes you, you'll probably thrive here. If it doesn't, it's better to know that upfront.

Come build the intelligence layer for global travel