3 months ago
Responsibilities
- Design, train, and evaluate computer vision and 3D ML models for extracting geometry and features from LiDAR and imagery.
- Lead ML research, prototype new approaches, run experiments, and determine what is ready to ship.
- Own models across problem framing, data strategy, training, evaluation, and integration into Digital Surveyor.
- Develop evaluation methodologies and metrics reflecting surveying and engineering accuracy requirements.
- Collaborate with ML infrastructure engineers to scale model training and inference.
- Partner with product teams to align model behavior with user needs and deliver production capabilities.
Requirements
- Master's or PhD in Machine Learning, Computer Vision, Computer Science, or a related field, or equivalent industry experience.
- Strong computer vision and deep learning foundation with hands-on experience training segmentation, detection, or 3D understanding models.
- Experience taking ML models from research or prototypes into production.
- Working knowledge of coordinate systems and 3D transforms.
- Proficiency in Python and a production-quality ML library such as PyTorch, JAX, or TensorFlow.
- Strong communication and collaboration skills with researchers, engineers, and product stakeholders.
- Bonus: experience with 3D deep learning architectures, point cloud backbones such as PTv3, sparse convolutions, or 3D detection and segmentation networks.
- Bonus: experience with large-scale imagery and 3D point cloud datasets.
- Bonus: experience with quantization, pruning, distillation, TensorRT, or ONNX Runtime for production model optimization.
- Bonus: familiarity with multi-GPU training and experiment management tools such as Weights & Biases.
- Bonus: publications or strong open-source contributions in computer vision or 3D machine learning.
Categories
About Mach9
Mach9 builds software that transforms reality capture datasets into high-precision 3D maps, accelerating how teams model and understand the world. Its flagship product, Digital Surveyor, automatically extracts features like utility poles, signs, curbs, striping and more from LiDAR and imagery datasets to create engineering-ready CAD and GIS deliverables. Surveying, engineering, and GIS teams work with Mach9 to reduce the time and cost of 3D mapping and deliver projects faster than ever before. Founded in 2021 and based in San Francisco, Mach9 is backed by Quiet Capital, Y Combinator, Soma Capital, Tiger Global, and Overmatch Ventures and trusted by leading engineering, construction and infrastructure organizations.