Research Scientist (Embodied AI)

Location: New York, United States; on-site

Salary: $170,000–$210,000 per year, plus equity

Company

hirify.global builds data infrastructure and proprietary real-world datasets for robotics and embodied AI.

What you will do

• Design, implement, and train foundation models for 3D hand pose estimation, dense mesh recovery, and kinematic tracking.
• Scale multi-view and temporal architectures across multi-GPU clusters and massive multimodal datasets.
• Develop loss functions and model architectures incorporating biomechanical constraints, temporal smoothness, and physical plausibility.
• Build egocentric hand-object interaction models that address motion blur, self-occlusion, rapid rotations, object tracking, tool segmentation, and contact modeling.
• Prototype perception models for tactile-visual fusion, action segmentation, and new hardware sensors.
• Integrate foundational tracking outputs into optimized pipelines for contact surfaces, object affordances, and robotic control policies.

Requirements

• Deep expertise in deep learning, 3D computer vision, articulated tracking, and hand pose estimation.
• Experience training large-scale vision models from scratch.
• Strong knowledge of parametric hand models, inverse kinematics, and dense mesh estimation.
• Mastery of PyTorch and deep learning scaling frameworks.
• Experience curating massive, multi-terabyte image and video datasets.
• Ability to work in a fast-paced R&D environment with rapidly changing priorities.

Nice to have

• First-author publications at CVPR, ICCV, ECCV, NeurIPS, or comparable venues.
• Experience with egocentric manipulation datasets such as Ego4D, Ego-Exo4D, DexYCB, or Epic-Kitchens.
• Experience writing custom CUDA kernels, differentiable mesh rendering, or collision and contact computation.

Culture & Benefits

• Access to proprietary, high-quality spatial and temporal ground-truth manipulation data.
• Blank-slate R&D with ownership of model development rather than legacy-system maintenance.
• Compute resources for large-scale state-of-the-art model training.
• Direct impact on embodied AI agents and robotic manipulation capabilities.
• Equity is included in the compensation package.