3D Reconstruction/Developer

About the Role
  
Our Client is dedicated to empowering humanity to design and build the next generation of infrastructure, both on Earth and beyond. They aim to enhance human potential by enabling people from all backgrounds, technical or otherwise, to easily design, forecast costs, and plan schedules for complex construction projects.
In this role, you will contribute directly to our 3D reconstruction initiatives, including real-time procedural 3D generation and computer vision model development. You will collaborate closely with the core engineering team and leadership.
They are looking for people who connect with their vision and are passionate about solving tough challenges, those who are not afraid to experiment, take bold risks, and learn from failure along the way.
  
Key Responsibilities

• Design and refine advanced computer vision and deep learning methods for object detection, classification, and segmentation.
• Research and prototype algorithms for procedural 3D reconstruction using 2D raster and vector data.
• Plan and conduct exploratory research to push the boundaries of machine perception and contextual image understanding.
• Develop innovative techniques for real-time 3D scene analysis and comprehension.

Requirements

• 5+ years of experience, or equivalent hands-on work in technically challenging projects within computer vision and 3D generation or design.
• Proficiency in Python and/or C++, with practical experience using leading computer vision and deep learning libraries (e.g., OpenCV, TensorFlow, PyTorch).
• Solid foundation in image analysis, machine learning, information theory, and 3D geometry.
• Familiarity with modern 3D tools, frameworks, and architectural concepts.
• Strong understanding of deep learning models and architectures, including Graph Neural Networks (GNNs) and Recurrent Neural Networks (RNNs).
• Skilled in data preprocessing, augmentation, and annotation for computer vision workflows.
• Experience deploying computer vision models and pipelines in real-world or production environments.
• Above all, a genuine passion for 3D technologies and a drive to learn quickly.

Preferred Qualifications

• Background or hands-on experience in structural, architectural, mechanical, or electrical engineering.
• Familiarity with Neural Radiance Fields (NeRFs), Building Information Modeling (BIM), and 3D reconstruction techniques.
• Proven experience addressing challenges in 3D scene and object reconstruction, as well as object detection.
• Experience with parametric or procedural mesh generation workflows.
• Practical knowledge of diffusion models, computer vision, or neural network architectures.