Machine Learning Engineer — Aerial Image Classification

Role overview Develop computer vision models that automatically identify assets, classify defects, detect changes, and estimate quantities in aerial inspection imagery. You will own the model lifecycle, from preparing data and training through evaluation and deployment, helping expand coverage across asset types, industries, and regions.

Responsibilities - Train, evaluate, and deploy detection, classification, and segmentation models for aerial inspection imagery, including images of roofs, solar arrays, transmission equipment, pipelines, and flare stacks. - Curate datasets and design augmentation strategies for challenges such as scale variation, nadir and oblique views, and combined thermal and RGB inputs. - Build training infrastructure and deploy models to cluster and edge environments. - Create evaluation tools that track regressions by class, region, and sensor. - Develop Python tooling across annotation, training, and model export workflows. - Work with the edge team to prune and quantize models for on-device inference.

Requirements - At least four years of experience building and shipping computer vision models in production. - Deep practical experience with PyTorch and modern detection methods and tools. - Strong Python engineering skills for maintainable production code. - Experience with detection or segmentation; overhead or industrial imagery is especially relevant.

Nice to have - Experience optimizing models for deployment with TensorRT, ONNX Runtime, or quantization.