У роли четкий исследовательский фокус, доступ к уникальным данным для робототехники и современный AI-стек, но указанная зарплата заметно ниже уровня похожих вакансий.
Ниже рынка
Современный стек
Четкая роль
Робототехника и AI
Доступ к данным
Описание вакансии
Текст:
Hirify AI/Оригинал
TL;DR
Research Scientist (3D Reconstruction) (Robotics/Embodied AI): Architecting and training proprietary 3D reconstruction pipelines that transform multimodal real-world data into photorealistic digital environments with an accent on neural rendering, dense 3D geometry extraction, and sim-to-real integration. Focus on scaling multi-GPU vision models, optimizing large scene representations for interactive client viewing, and integrating spatial assets into robotics perception pipelines.
Location: Hybrid in the Toronto GTA or New York
Compensation: CA$200K–CA$230K total wage range, plus potential equity
Company
hirify.global builds data infrastructure for robotics and embodied AI, including systems for data capture, labeling, and hardware-enabled workflows used to train and validate humanoid and embodied AI systems.
What you will do
Architect, train, and scale neural reconstruction models for dense 3D geometry extraction across multi-GPU clusters.
Build high-fidelity digital twins using techniques such as 3D Gaussian Splatting and NeRFs.
Integrate reconstructed environments into robotics perception pipelines, world models, and sim-to-real training.
Compress, optimize, and package large 3D scene representations for high-framerate client-facing viewers.
Prototype methods for separating dynamic actors from static environments and integrating new hardware sensors.
Requirements
Deep expertise in 3D computer vision, neural rendering, multi-view geometry, and spatial modeling.
Experience training and scaling large 3D vision models from scratch with PyTorch and multi-terabyte datasets.
Experience writing custom CUDA kernels for 3D operations, ray marching, or rasterization.
Ability to work in a hybrid role based in the Toronto GTA or New York.
Nice to have
First-author publications at CVPR, ICCV, ECCV, or similar top-tier venues in 3D deep learning, neural rendering, or spatial transformers.
Public projects involving WebGL, custom rendering engines, or optimization of splats for interactive viewing.
Culture & Benefits
Access to proprietary spatial and temporal ground-truth data at significant scale.
Opportunity to develop spatial models and environmental reconstructions for embodied AI systems.
Equity may be included in the total rewards package.
Inclusive hiring practices and interview accommodations are available.