1. Data, Video & Annotation Management

• Collaborate with the validation team to understand and review annotations.

• Extract datasets directly from Postgres using SQL and analyse data distributions.

• Clean, preprocess, and structure image and video datasets.

• Implement video-quality checks (blur, brightness, orientation, motion, duration).

• Perform dataset balancing, augmentation, and versioning.

2. Model Development

• Build and optimise deep learning models for object detection, classification, and counting.

• Conduct experiments with proper tracking (MLflow, W&B) and reproducibility.

• Evaluate models using mAP, precision, recall, and F1-score.

• Ensure robustness across diverse image/video conditions.

• Experience with architectures such as CNNs and transformer-based models (e.g., DETR, ViT).

3. Inference & Deployment

• Develop efficient inference pipelines for cloud, edge, or mobile environments.

• Optimise models using ONNX, TensorRT, quantisation, or similar tools.

• Package and deploy models using Docker, GitHub Actions, and AWS (EKS, Lambda, S3).

• Monitor production performance and iterate based on runtime metrics.

4. Collaboration & Documentation

• Work with software engineers to integrate ML outputs into applications.

• Coordinate with product and validation teams to refine accuracy requirements.

• Maintain documentation on datasets, model versions, training processes, and deployments.

• Contribute to CI/CD workflows for ML components.

Required Skills

• Minimum 5 years of hands-on experience in Machine Learning, including at least 3 years focused on computer vision or deep learning.

• Bachelor’s degree in Computer Science, Engineering, or a related field; Master’s degree preferred.

• Strong Python engineering skills (data pipelines, automation, tooling).

• Proficiency with PyTorch or TensorFlow.

• Solid experience in computer vision and video processing (OpenCV, ffmpeg).

• Strong SQL skills and experience with Postgres or similar databases.

• Familiarity with Docker, Git, and CI/CD (GitHub Actions).

• Experience deploying ML models in production (AWS preferred).

• Strong analytical and problem-solving abilities.

Nice to Have

• Experience with biological or diagnostic imaging datasets.

• Knowledge of classical CV algorithms.

• Familiarity with MLOps tools (MLflow, DVC, W&B).

• Experience working with frame-level video analysis or optical flow.

• Familiarity with transformer-based computer vision models (DETR, ViT, Swin Transformer).

This is a great opportunity for someone that wants to own a project with high impact. You will be given autonomy and responsibility for an exciting development in a fast-paced and innovative team working at the cutting edge of technology, AI and veterinary medicine.