Stay safe — Scam Alert
Job scams are on the rise — some use our name or logo to steal your personal or financial information. We never send links via WhatsApp, Telegram or SMS. We post only on www.jobnomix.com. Never send money to companies — jobs on JobNomix will never require payment from applicants.
About LatentWorlds
LatentWorlds builds the data infrastructure layer for physical intelligence workflows. Our data platform helps robotics and autonomy teams capture, move, store, query and curate multimodal sensor data across cloud and on-premise environments. We connect this work to robot learning and world modeling, with a focus on reliable real-world deployment.
The role
Work on a focused research problem at the intersection of physical data and intelligent behaviour. Depending on your strengths, your project will centre on world modeling, hand motion tracking, video captioning or data annotation pipelines. You are not expected to be expert in all four. You will work with a research mentor, build reproducible experiments and connect your findings to real data and practical workflows.
What you'll do
- With your mentor, define a research question, reproduce relevant baselines and choose evaluation criteria before scaling experiments.
- Explore world models that learn temporal structure and action-conditioned dynamics from multimodal trajectories.
- Develop or evaluate hand motion tracking and reconstruction, including temporal consistency, occlusion and interactions with objects.
- Improve grounded video captioning, temporal segmentation or annotation pipelines, including uncertainty-aware quality checks and human review.
- Build clean datasets and evaluation splits; run ablations and analyse failure cases rather than relying on aggregate scores alone.
- Contribute tested research code, experiment records and technical reports; support publications or open-source releases where appropriate.
What you bring
- Current MSc or PhD study in machine learning, robotics, computer vision or a related field, or equivalent research experience.
- Strong Python and practical experience with a deep-learning framework such as PyTorch; comfortable training and evaluating models.
- Depth in at least one relevant area: generative or world models, 3D vision and tracking, vision-language models, or data-centric ML.
- Careful experimental reasoning, clear technical English and evidence of sustained research or engineering work through projects, code or papers.
Nice to have
Experience with egocentric video, geometry, multimodal temporal data, scalable annotation or robotic manipulation. Publications are welcome, not required.
Why join
Own a substantive project with close mentorship and a direct path from research to infrastructure or model improvements. Paid placement; scope, dates and duration agreed with you.
How to apply
Email careers@latentworlds.ai with the role title, your CV, code or research links, preferred research area, and availability and internship duration.
LatentWorlds AI is an equal opportunity employer.
