Role Purpose
We’re looking for a Staff/Senior Machine Learning Engineer with deep expertise in computer vision and biometrics to lead the design and scaling of face recognition systems in production. You’ll build and train models, and own ML systems end-to-end on AWS. The final job level for this role will be determined following the interview process.
What You’ll Do
- Lead the design and development of computer vision systems for biometrics (face attributes, detection, quality, and recognition).
- Conduct rigorous fairness analysis and benchmarking of biometric models across various datasets and operating conditions.
- Architect, train, and optimize models using PyTorch, Tensorflow, and/or JAX.
- Own and evolve end-to-end ML pipelines, from data ingestion to deployment. Design automated pipelines (Airflow) for data ingestion and cleaning, curate balanced training sets, and generate synthetic data to address quality and diversity gaps.
- Optimize models for low-latency inference (quantization, distillation, TensorRT/ONNX) and manage deployment on AWS.
- Mentor ML engineers, conduct code/design reviews, and drive technical best practices across the Computer Vision team.
What We’re Looking For
- 5+ years of industry experience in Machine Learning, with at least 3 years dedicated to Biometrics or Face Analysis.
- Deep expertise in computer vision and biometrics, especially face recognition.
- Understanding of algorithmic bias in Computer Vision and practical experience measuring and mitigating disparate impact.
- Expert proficiency in Python and vision libraries such as Pillow, OpenCV, PyTorch, etc. Ability to write clean, modular, production-ready code.
- Experience designing end-to-end ML pipelines (Data to Train to Deploy) and working with workflow orchestrators like Airflow.
- Hands‑on experience scaling training jobs on multi‑GPU clusters and deploying services on AWS (SageMaker, EC2, EKS).
Nice to Have
- Research publications in CVPR, ICCV, ECCV, or FG related to face recognition, image quality assessment, or fairness.
- Experience with vector databases (e.g., Milvus, Faiss) and approximate nearest neighbor (ANN) search algorithms.
- Familiarity with privacy, security, and compliance in biometric systems.
- Experience porting models to edge or mobile devices using CoreML, LiteRT, and/or TFLite.
- Experience using GANs or diffusion models to generate synthetic faces for training.
- Strong communication skills.
Equal Opportunities
Jumio is a collaboration of people with different ideas, strengths, interests and cultures. We welcome applications and colleagues from all backgrounds and of all statuses.