At a Glance
- Tasks: Lead original research in multimodal data curation and prototype cutting-edge AI workflows.
- Company: Join a pioneering company backed by top-tier venture funds, shaping the future of Physical AI.
- Benefits: Competitive salary, access to proprietary data, and opportunities to publish your research.
- Other info: Work on-site in Central London with excellent career growth and networking opportunities.
- Why this job: Make a real impact in robotics and AI while working with leading experts in the field.
- Qualifications: PhD or equivalent experience in computer vision or machine learning with a strong research background.
The predicted salary is between 54000 - 66000 Β£ per year.
This role is with one of Dex's trusted partner companies.
We work closely with their teams to truly understand their culture, goals, and what they're looking for, so we can match you with the right opportunity and give you context about the role before you commit to a process.
Dex is an AI recruiter agent that helps you run your job search.
Tell Dex your stack, seniority, and what you want to build.
We will manage your applications and surface other opportunities that are a fit.
The role
Frontier robotics and embodied AI live or die by the quality of their training data, yet the infrastructure to curate, evaluate, and scale multimodal real-world inputs remains largely broken.
Backed by over $100M from top-tier venture funds and serving more than 300 AI and robotics teams globally, this company builds the foundational data layer for Physical AI.
With continuous access to massive, proprietary multimodal streams, their dedicated frontier research lab operates on real-world edge cases that academic benchmarks cannot replicate.
As a Research Engineer, you will lead original research directly driven by open problems from leading robotics labs.
This is not ivory-tower benchmarking on synthetic toys, nor is it standard ML operations; you will formulate hypotheses, run experiments end to end, build working prototypes, and ship your methods straight into a core platform while retaining the mandate to publish.
Your scope covers agentic multimodal workflows, automated data curation, and pushing the boundary on egocentric video and 3D scene understanding.
- The work
- Formulate hypotheses, design rigorous experiments, and drive original research targeting open challenges in multimodal data curation, model evaluation, and automated quality control.
- Prototype agentic workflows that leverage modern vision-language models to interpret complex real-world ontologies and unstructured sensor inputs.
- Tackle frontier video problems, advancing models from egocentric dense captioning toward deep 3D spatial and temporal scene understanding.
- Build end-to-end prototypes in Py Torch and partner with ML systems engineers to transition proven architectures directly into production.
- Publish novel methodologies and represent the research lab at major computer vision and machine learning conferences.
What You Bring
- Ph D in computer vision, machine learning, or a related discipline, or demonstrable equivalent experience driving applied research.
- Demonstrated track record of owning original research contributions, evidenced by peer-reviewed publications, a doctoral thesis, or high-impact open-source multimodal ML projects.
- Deep technical depth in computer vision and multimodal architectures-spanning vision-language models, representation learning, and video or 3D understanding-backed by rigorous statistical foundations.
- Production-grade Python and Py Torch skills with direct experience running large-scale training and evaluation experiments end to end.
- Ability to work on-site four days a week in Central London.
- Why apply through Dex
Frontier Physical AI roles are rare, highly contested, and often hard to differentiate from early-stage hype that lacks real data.
Dex cuts through the noise, matching your research track record directly to the hiring leadership without the friction of cold applications.
You get fully briefed on the lab's proprietary data pipelines, compute setup, and technical roadmap before stepping into the process.
As part of the recruitment process at Dex, we process your personal data in accordance with our Privacy Notice for Job Applicants.
This notice explains how and why your data is collected and used, and how you can contact us if you have any concerns.
Research Engineer employer: Dex
At Dex, we pride ourselves on being an exceptional employer, offering a dynamic work culture that fosters innovation and collaboration. Our London-based team enjoys a hybrid working model, competitive benefits, and ample opportunities for professional growth, all while contributing to cutting-edge technology in the healthcare sector. Join us to make a meaningful impact in real-time systems and edge deployment, surrounded by a supportive team dedicated to excellence.