> Markdown version of [/jobs/ext/2356493-senior-computer-vision-engineer](https://www.wearedevelopers.com/jobs/ext/2356493-senior-computer-vision-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Senior Computer Vision Engineer - **Company:** Avalon Artificial Intelligence Limited - **Location:** UK (Remote available) - **Experience:** Expert - **Salary:** £26,000.0 - £27,000.0 - **Contract:** Permanent contract - **Skills:** Artificial Intelligence, Automatic Number Plate Recognition, Computer Vision, Nvidia CUDA, Data Files, Linux, Python (Programming Language), Machine Learning, Object Detection, OpenCV, Tensorflow, Graphics Processing Unit (GPU), Pytorch, Deep Learning, Model Validation, Git, ONNX (Open Neural Network Exchange) Format, Machine Learning Operations, TensorRT, Docker - **Published:** August 25, 2026 - **Apply:** https://uk.indeed.com/viewjob?jk=ea73d9b484a89951 ## About the Role We are looking for someone with practical experience in computer vision, machine learning or deep learning, preferably with experience taking models beyond experimentation and into real-world applications. You should have experience with some of the following: * Python * PyTorch or TensorFlow * OpenCV * Object detection models such as YOLO, RT-DETR, Faster R-CNN or similar * Image classification and segmentation * Multi-object tracking * Deep learning model training and fine-tuning * Dataset preparation and augmentation * Computer vision evaluation metrics * GPU-based model training and inference * ONNX or TensorRT * Linux * Git * Docker Experience with the following would be particularly beneficial: * CCTV or surveillance computer vision * Real-time video analytics * Face recognition * Automatic Number Plate Recognition (ANPR/ALPR) * Person re-identification * Vehicle re-identification * Multi-camera tracking * Behaviour or activity recognition * Vision Transformers * Video understanding models * Edge AI * NVIDIA GPUs and CUDA * Model quantisation and inference optimisation * Production ML/MLOps systems We do not expect candidates to have experience with every technology listed. The Ideal Candidate You will be someone who: * Enjoys solving difficult real-world computer vision problems rather than working only with clean benchmark datasets. * Understands that a model performing well in a laboratory does not automatically mean it will perform reliably in production. * Can analyse false positives, false negatives and difficult edge cases and determine why they occur. * Thinks carefully about dataset quality and understands how strongly data affects model performance. * Can evaluate models objectively using measurable evidence. * Is comfortable experimenting with new architectures and research techniques. * Can balance model accuracy with inference speed and computational requirements. * Understands the challenges of video analytics including occlusion, lighting variation, camera movement, perspective and low-quality imagery. * Has a strong engineering mindset and can turn research ideas into maintainable production systems. * Can work independently while collaborating closely with software, platform and product teams. ## Description This is a hands-on AI engineering role focused on building computer vision systems that operate reliably in real-world environments. You will work on detection, tracking, recognition, behaviour understanding and visual intelligence models used across security, monitoring and automation applications. You will work closely with our AI, software and platform engineering teams to take models from experimentation through validation and into production customer deployments. Key Responsibilities * Design, train, evaluate and improve computer vision and deep learning models for security applications. * Develop models for applications such as: * Person and vehicle detection * Object detection and classification * Multi-object tracking * Face detection and recognition * Vehicle and number plate recognition * Intrusion and perimeter monitoring * Anomaly and unusual-behaviour detection * Activity and event recognition * Occupancy and movement analysis * Visual scene understanding * Develop robust models that work across different cameras, lighting conditions, environments and viewing angles. * Improve model accuracy, precision, recall and reliability while reducing false positives and false negatives. * Build and curate datasets for training, validation and benchmarking. * Design data annotation and dataset-quality processes. * Fine-tune and optimise existing models for Avalon-specific use cases. * Research and evaluate new computer vision architectures, models and techniques. * Develop multi-camera and temporal computer vision approaches where required. * Improve object identity and tracking across frames and camera events. * Develop algorithms for combining information across multiple frames rather than relying only on individual images. * Evaluate model performance using real-world customer footage and representative test datasets. * Investigate difficult edge cases and systematically improve model behaviour. * Optimise models for GPU, edge and production inference environments. * Improve inference speed, memory utilisation and computational efficiency. * Work with technologies such as ONNX, TensorRT or similar inference optimisation frameworks where appropriate. * Integrate trained models into production applications and computer vision pipelines. * Work with Platform Engineers to deploy and monitor models in customer environments. * Create automated model evaluation and regression-testing processes. * Monitor production model performance and identify model drift or performance degradation. * Document experiments, datasets, model versions and evaluation results. * Contribute to the architecture and technical direction of Avalon's computer vision platform. ## Related Videos - [Docker network without Docker](https://www.wearedevelopers.com/videos/1418-docker-network-without-docker) - [How a Small Team Shrank a Microsoft Monorepo by 94%](https://www.wearedevelopers.com/videos/1236-how-a-small-team-shrank-a-microsoft-monorepo-by-94) - [Docker Compose: Rediscovered](https://www.wearedevelopers.com/videos/1978-docker-compose-rediscovered) - [How computers learn to see – Applying AI to industry](https://www.wearedevelopers.com/videos/756-how-computers-learn-to-see-applying-ai-to-industry) - [Docker exec without Docker](https://www.wearedevelopers.com/videos/1094-docker-exec-without-docker) - [Docker build without Docker](https://www.wearedevelopers.com/videos/100114-docker-build-without-docker) ## Related Articles - [How to Become an AI Engineer](https://www.wearedevelopers.com/magazine/331-how-to-become-an-ai-engineer) - [Dev Digest 121 - AI goes offline](https://www.wearedevelopers.com/magazine/456-dev-digest-121-ai-goes-offline) - [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development) - [Coffee with Developers - Maria Apazoglou - Making AI understandable for all in production](https://www.wearedevelopers.com/magazine/475-coffee-with-developers-maria-apazoglou-making-ai-understandable-for-all-in-production) - [Dev Digest 137 - AI'm not sure about this](https://www.wearedevelopers.com/magazine/485-dev-digest-137-ai-m-not-sure-about-this) - [Navigating the AI Shift](https://www.wearedevelopers.com/magazine/629-navigating-the-ai-shift)