Design and train modern computer vision models (CNNs, Vision Transformers, foundation models) to solve novel perception tasks
Build and maintain scalable training pipelines using PyTorch and HPC infrastructure (e.g., Slurm, distributed training)
Develop data curation and active learning workflows
Optimize models for deployment (ONNX, TensorRT, containerization)
Test and validate models in both cloud and edge environments
Build reproducible experimentation workflows (version control, experiment tracking, configuration management)
Drive experimental cycles: define hypotheses, implement techniques from literature, evaluate results, and present recommendations
Translate research ideas into production-ready implementations
Design, implement, and test ML-related components and supporting software
Perform statistical analysis and fine-tune model performance based on production and field feedback

3+ years of full-time professional experience in computer vision-focused machine learning (not student internship)
Strong PyTorch experience, including custom layers, loss functions, datasets, dataloaders, and training loops
Experience with modern vision architectures (CNNs, ViTs, DETR-style models, foundation models)
Experience building or contributing to production ML systems
Solid software engineering fundamentals (testing, version control, clean code principles)
Strong communication skills and ability to explain complex technical topics clearly
Engineering degree or equivalent practical experience

Machine Learning Engineer

Key skills