Job DescriptionAbout the Job: We are seeking a seasoned
Machine Learning Engineer - Computer Vision to design, optimise, and deploy deep learning models for large-scale, real-time edge inference. In this role, you will work on the end-to-end lifecycle of computer vision models-from training and evaluation to optimisation, automated governance, and edge deployment-while advancing MLOps capabilities on Google Cloud. You will work at the intersection of deep learning, cloud infrastructure, and edge AI, building reliable, high-performance solutions that scale across devices and continuously improve through automation and data driven evaluation.
Office Location: Toronto
Employment Type: Permanent
Role Type: New position - current requirement
Work Arrangement: Hybrid (2 days in office per week)
Position Responsibilities:- Computer Vision Development: Design, train,evaluate, and fine-tune state-of-the-art deep learning models for imageclassification and object detection tasks.
- Pipeline Enhancement: Maintain, optimize and addadvanced MLOps capabilities to existing Vertex AI Kubeflow Pipelines(KFP).
- Model Optimization & Conversion: Manage thecomplex conversion of models from frameworks like TensorFlow into highlyoptimized TensorFlow Lite (TFLite) artifacts for edge inference (e.g.,handling Int8 full integer quantization and hardware-specific acceleration).
- Edge Artifact Management: Architect the deploymentflow to save optimized edge models to Google Cloud Storage (GCS) andmanage model versioning for seamless edge-device retrieval, bypassingtraditional Vertex AI Endpoints.
- Automation & Reliability: Implement automatedevaluation gates to ensure newly trained models outperform existingproduction models before edge deployment.
RequirementsRequired Qualifications:- Experience: 3- 6 years in Machine LearningEngineering, preferably Computer Vision.
- Deep Learning Foundation: Strong mathematical andarchitectural understanding of deep learning concepts, specificallyConvolutional Neural Networks (CNNs) and standard object detectionarchitectures.
- Framework Mastery: Deep, hands-on expertise withTensorFlow 2.x and/or PyTorch.
- Edge ML: Proven experience optimizing deep learningmodels for edge devices using TFLite (e.g., post-training quantization,pruning, handling custom ops).
- GCP MLOps: Strong proficiency in Google CloudPlatform, specifically building and running custom components in Vertex AIPipelines (KFP).
- Programming: Advanced programming skills in Python,with experience containerizing ML workloads using Docker.
- Cloud Infrastructure: Solid understanding of GoogleCloud Storage (GCS) for managing massive datasets and handling modelartifact hand-offs.
- Critical thinking, Effective communication skills -verbal and written, Problem solving, and Dealing with complexity
Preferred Qualifications:- YOLO Expertise: Hands-on experience with theUltralytics YOLOv8 ecosystem, specifically bridging PyTorch YOLO weightsto TensorFlow/TFLite edge deployments.
- Data Orchestration: Experience using Google CloudComposer (Apache Airflow) to schedule and trigger complex ML trainingpipelines based on data arrival or model drift.
- Scalable Data Processing: Familiarity with GoogleCloud Dataflow (Apache Beam) for large-scale, parallelized imagepreprocessing, augmentation, and dataset formatting (e.g., generatingTFRecords).
- CI/CD for ML: Experience with continuousintegration and continuous deployment practices specifically tailored formachine learning models.
- Generative AI: Knowledge or experience inGenerative AI architectures, with experience building Retrieval-AugmentedGeneration (RAG) pipelines and developing multi-agent systems.
BenefitsSalary Range: CAD $100,000 - $110,000/ year
The final compensation offered will depend on local market conditions and geographic location, as well as job-related factors such as the candidate's knowledge, skills, qualifications, relevant experience, and education/training. Compensation may also include additional components such as benefits, and/or other incentives, where applicable. In accordance with new employment standards requirements, we retain copies of this job posting and applicant information for three (3) years after the posting is removed. We do not use AI technology; all applications are also reviewed by our recruitment team.
