Responsibilities :

- Design, develop, and deploy production-grade face recognition and biometric verification systems at enterprise scale.

- Build, train, and optimise deep learning models for face detection, recognition, embedding generation, and biometric verification with a focus on high accuracy, fairness, robustness, and low-latency inference.

- Architect and own end-to-end ML pipelines covering data preparation, model training, experimentation, deployment, monitoring, and continuous model improvement.

- Develop scalable model serving solutions using PyTorch, TorchServe, Docker, Kubernetes, and AWS cloud infrastructure.

- Lead technical design discussions, define ML architecture, conduct design reviews, and mentor engineers on machine learning best practices.

- Collaborate with AI researchers, platform engineers, and product teams to deliver secure, scalable, and production-ready computer vision solutions.

- Drive continuous improvements in model performance, bias mitigation, system reliability, and production scalability.

Requirements :

- 5+ years of Machine Learning experience with 4+ years specialising in Face Recognition, Biometrics, Face Analysis, or Computer Vision.

- Strong expertise in PyTorch and production-grade deep learning systems.

- Hands-on experience with face recognition frameworks such as ArcFace, FaceNet, InsightFace, RetinaFace, MTCNN, or similar technologies.

- Experience deploying large-scale ML models using TorchServe, Docker, Kubernetes, and AWS.

- Experience with MLflow or Weights & Biases (W&B) for experiment tracking and model lifecycle management.

- Strong Python programming skills with expertise in scalable software design and production ML architecture.

- Experience building highly available, low-latency inference systems for large-scale production environments.

- Strong understanding of distributed training, model optimisation, and GPU acceleration.

- Proven experience leading technical architecture discussions, mentoring engineers, and driving engineering excellence.

- Excellent problem-solving, communication, and stakeholder management skills.

Preferred Skills :

- Experience with Vision Transformers (ViT), ResNet, MobileNet, YOLO, or similar computer vision architectures.

- Knowledge of TensorRT, ONNX Runtime, CUDA, or other model optimisation frameworks.

- Experience in fairness, bias mitigation, and responsible AI for biometric systems.

- Familiarity with SageMaker, Kubeflow, Triton Inference Server, or distributed ML platforms