Senior Machine Learning Engineer (Computer Vision – Medical Device)

Job Type: Full-time, Permanent

Location: Richmond Hill (On-site)

Experience Level: Senior

About the Role

At LightArt Labs, we are developing a medical device that relies on a production-grade computer vision image classification pipeline. We have already completed model exploration and selected a strong baseline architecture. The remaining work is to close the gap to production readiness by improving model performance, robustness, and efficiency, while tightening evaluation discipline and delivery quality.

This is a critical senior role. You will collaborate closely with software engineering and our medical/biological team, and you must be able to deliver results under tight deadlines.

Responsibilities

• Own the end-to-end cycle for training, evaluation, iteration, and release readiness of a transformer-based vision classification system.
• Improve model accuracy, robustness, and generalization beyond an existing strong baseline (not greenfield exploration).
• Improve efficiency (latency, memory footprint, loading time, throughput) while preserving accuracy.
• Build and maintain a high-integrity evaluation framework: realistic dataset splitting strategy appropriate for real-world bias/leakage risks; per-class metrics, calibration, confusion matrices; disciplined ablations and reproducible experiments.
• Design and maintain image pre-processing that is stable, versioned, testable, and production-safe (normalization, resizing/cropping, channel handling, illumination variability handling, artifact removal as appropriate).
• Drive deep error analysis and targeted fixes: failure modes, label issues, data gaps, and measurable improvements.
• Collaborate with software to ensure models are exportable and deterministic (stable preprocessing, versioning, reproducible outputs).

Must-Have Qualifications

• 5+ years hands-on experience in ML/CV with a track record of shipping models into production or production-adjacent environments.
• Bachelor’s degree in Computer Science, Engineering, or related field, and 3+ year of experience in Software Engineering, Systems Engineering, or related area.
• Strong Python skills with TensorFlow/Keras and/or PyTorch.
• Transformer-based vision models experience is required (e.g., ViT/Swin/CvT/MaxViT or similar): training, fine-tuning, evaluation, and practical trade-offs.
• Image pre-processing experience is required: designing and owning the preprocessing pipeline, not just calling a standard transform stack.
• Proven ability to improve an existing model through systematic iteration: augmentation strategy, class imbalance handling, loss/optimization tuning, and controlled ablations.
• Strong evaluation fundamentals and the ability to avoid common pitfalls (data leakage, metric gaming, overfitting to validation).
• Demonstrated ability to execute under tight timelines with high accountability.

Strong Assets (Nice to Have)

• Medical device / diagnostics / regulated product exposure.
• Comfortable collaborating with biology/medical teams and translating domain reality into ML requirements.
• Model export and deployment awareness: ONNX, inference optimization, or production constraints.
• Working knowledge of C# or prior integration into a C# application.
• Self-supervised learning experience (BYOL-style or similar).

How to Apply

Please include: resume or LinkedIn, plus brief descriptions of 1–2 relevant projects (dataset scale, model used, what improved, and how you validated).

We are committed to creating an inclusive, accessible work environment where all individuals feel valued, respected, and supported. In accordance with the Accessibility for Ontarians with Disabilities Act (AODA), we will provide accommodations throughout the recruitment, selection, and employment process for applicants with disabilities. If you require an accommodation at any stage of the hiring process, please let us know at hr@lightartlabs.com, and we will work with you to meet your needs.

Job Types: Full-time, Permanent

Pay: $110,000.00-$160,000.00 per year

Benefits

• Dental care
• Disability insurance
• Employee assistance program
• Extended health care
• Life insurance
• Paid time off
• Vision care

Application question(s):

• Do you have hands-on experience with vision transformers in real training/evaluation work?
• Do you have a concrete case where you improved a model beyond a baseline using disciplined experiments and error analysis?
• Do you have hands-on experience with owning reliability, reproducibility, or production constraints?

Work Location: In person