Job Specifications
Junior Applied Scientist
London — Onsite
Our client are building context-aware agentic systems that augment data science, software, and product teams inside large enterprises. The systems reason over complex organisational data, take multi-step actions, and integrate deeply with existing business workflows. We are now expanding the Applied Science team and are looking for exceptional early-career researchers who want to work on cutting-edge ML in a fast-moving product environment.
This role is ideal for candidates finishing (or recently completing) a PhD at a top university, with a research track record at leading ML/AI conferences.
As this is a role within the Applied team, candidates must be confident in their ability to move research through to production.
What You’ll Work On
Develop and iterate on core models for context-aware enterprise agents, including planning, retrieval, grounding, and multi-step decision pipelines.
Conduct research on any ML frontier relevant to agentic systems e.g., LLMs, multimodal models, structured reasoning, data generation, or video-text modelling.
Build robust training, evaluation, and benchmarking setups for new model variants and agent behaviours.
Prototype new agent capabilities and work closely with engineering teams to turn them into production-quality features.
Analyse failure modes of agents in real customer environments and design methods to improve reliability, grounding, and actionability.
Contribute to internal research papers, technical memos, and (where appropriate) external publications.
Who You Are
Currently completing or recently completed a PhD in Physics, Mathematics, Machine Learning, Computer Science, or a related field from a top university.
Strong publication record at leading venues such as NeurIPS, ICML, ICLR, ACL, CVPR, ICCV, or EMNLP.
Solid understanding of modern ML architectures (transformers, diffusion, retrieval-augmented systems, reinforcement learning, etc.).
Strong coding skills in Python and experience with at least one major ML framework (PyTorch, JAX, TensorFlow).
Ability to bridge high-level research with practical implementation.
Curious, humble, and excited to work on hard technical problems with real enterprise impact.
Nice to Have
Experience in agentic or autonomous decision-making systems (not required).
Prior work on LLM alignment, grounding, tool use, or planning.
Experience with enterprise data environments, knowledge graphs, or retrieval systems.
Hands-on experience with large-scale training, distributed compute, or evaluation tooling.
Why Join?
Work on a high-impact ML problem space that blends deep research with real product deployment.
Collaborate with a world-class team of scientists and engineers with prior experience at leading AI labs and tech companies.
Opportunity to publish, attend conferences, and lead impactful research directions early in your career.
Competitive salary, equity, and benefits.
Located in central London with a strong in-person research culture.
About the Company
We believe that advanced technology is paving the way to a more sustainable future.
We believe in going the extra mile to provide our customers with a competitive edge.
Since our inception in 2010, Cubiq has established a reputation for excellence in the delivery of high quality engineering talent to the most ground-breaking technology companies in the world.
Cubiq methods are subtle, data-driven, and highly effective, which is how we have become widely recognised as leading experts within engineering and technology recr...
Know more