About the job
Jumio seeks a Senior Machine Learning Engineer with a strong background in computer vision and biometrics. This role centers on designing, building, and scaling face recognition systems for production use. Responsibilities include developing and training models, managing their lifecycle on AWS, and advancing the maturity of Jumio’s machine learning systems. Final seniority will be determined during the interview process.
Main responsibilities
- Design and develop advanced computer vision solutions for biometric applications, covering face attribute analysis, detection, quality assessment, and recognition.
- Conduct fairness assessments and benchmark biometric models across diverse datasets and operational scenarios.
- Build, train, and optimize machine learning models using frameworks such as PyTorch, TensorFlow, or JAX.
- Oversee end-to-end ML pipelines, from data ingestion through deployment. Use automated workflows (Airflow) for data collection and cleaning, curate balanced datasets, and generate synthetic data to address diversity or quality gaps.
- Optimize models for low-latency inference with techniques like quantization and distillation, and deploy using tools such as TensorRT and ONNX on AWS infrastructure.
- Mentor other ML engineers, conduct code and design reviews, and help shape best practices for the Computer Vision team.
Requirements
- Minimum 5 years of professional experience in machine learning, with at least 3 years focused on biometrics or face analysis.
- Deep understanding of computer vision and biometrics, particularly face recognition technologies.
- Awareness of fairness and ethical issues in AI, including algorithmic bias in computer vision and experience measuring or mitigating disparate impact.
- Advanced engineering skills in Python and libraries such as Pillow, OpenCV, and PyTorch, with the ability to write clean, modular, production-ready code.
- Experience designing ML pipelines from data collection to deployment and familiarity with workflow orchestrators like Airflow.
- Hands-on experience scaling training jobs on multi-GPU clusters and deploying services on AWS (SageMaker, EC2, EKS).
Preferred qualifications
- Research publications at conferences such as CVPR, ICCV, ECCV, or FG in areas like face recognition, image quality assessment, or fairness.
- Experience with large-scale search technologies, including vector databases (Milvus, Faiss) and approximate nearest neighbor (ANN) algorithms.
Location
This position is based in Montreal.
