Machine Learning Engineer

Title: Machine Learning / Computer Vision Engineer

Location: Waterloo, ON (onsite, occasional hybrid days)

Type: Full-time or part-time, 6-month to 1-year contract

About Neatco

Neatco Engineering builds AI vision and robotics systems for manufacturing and recycling. Our IntelSight platform uses deep learning and industrial imaging to run real-time quality inspection on production lines across automotive, manufacturing and recycling.

The role

You'll develop and improve the deep learning models behind IntelSight, from research ideas to models running on real factory floors. You'll work with real manufacturing datasets, industrial cameras and 3D sensors, and NVIDIA GPU and edge hardware.

This role suits people who enjoy applied research. We especially welcome master's and PhD students, postdoctoral researchers, and people who recently completed a graduate degree.

What you'll do

• Train, evaluate and optimize models for detection, instance and semantic segmentation, classification and anomaly detection.
• Improve IntelSight's accuracy, latency and robustness in industrial conditions.
• Explore continual, incremental and few-shot learning, so new defect types need minimal retraining.
• Build preprocessing, augmentation, dataset and evaluation pipelines.
• Deploy models on NVIDIA GPUs and edge devices with TensorRT and ONNX.
• Work with 2D cameras, 3D laser profilers, depth sensors and multispectral data.
• Document experiments and benchmarks clearly.

What you bring

• Strong Python and PyTorch.
• Hands-on experience building computer vision models, such as detection or segmentation.
• Solid grasp of training, validation, overfitting and evaluation.
• Comfortable with Linux, Git and GPU environments.
• Degree, or current studies, in CS, engineering, mechatronics or a related field.

Nice to have

• Detectron2, YOLO, U-Net or vision transformers.
• Continual learning, domain adaptation, self-supervised learning or anomaly detection.
• CUDA, TensorRT, ONNX Runtime or Jetson.
• 3D point clouds, depth maps or 16-bit industrial images.
• Publications, a thesis or open-source work in computer vision.

Details

• Full-time or part-time contract for 6 months to 1 year, with potential for an extension.
• Compensation is discussed at the first interview.

How to apply

Apply here: https://tally.so/r/WJ9v8R. It takes about 10 minutes and asks for your resume as a PDF.

Our process is a screening call, then an onsite technical assessment, then a final conversation with our CEO and CTO.

Neatco uses AI to help screen and assess applications. Every hiring decision is made by people.

Accommodations are available on request at any stage of the process.