.avif)
The table of content
About the Project
We are looking for a Computer Vision Engineer to join a team building visual inspection and quality monitoring solutions for retail and light industrial environments. The platform detects product placement issues, packaging defects, shelf compliance, object presence, and visual anomalies from images and short video streams.
The stack includes Python, PyTorch, OpenCV, Ultralytics YOLO or similar detection frameworks, data annotation workflows, model evaluation, Docker, and deployment to cloud or edge environments. The role requires regular communication with international product, data, and engineering teams, so strong spoken English is essential.
What You Will Do
- Develop and improve computer vision models for detection, classification, segmentation, and visual anomaly use cases.
- Work with image and video datasets, annotations, labelling quality, augmentation, and dataset versioning.
- Train, evaluate, and optimise models for accuracy, latency, robustness, and real-world edge cases.
- Build preprocessing and postprocessing pipelines using OpenCV and Python.
- Support deployment of models as APIs, batch inference jobs, or edge inference components.
- Collaborate with product teams to define acceptance criteria and measure model performance in business terms.
- Investigate production issues related to image quality, lighting, camera angle, false positives, and drift.
What We Are Looking For
- 3+ years of commercial experience in computer vision or applied machine learning.
- Strong Python skills and practical experience with PyTorch and OpenCV.
- Experience with object detection, classification, segmentation, or anomaly detection models.
- Understanding of dataset preparation, annotation quality, augmentation, metrics, and model validation.
- Experience deploying ML/CV models with Docker, APIs, batch jobs, or edge devices.
- Ability to analyse model errors and improve performance using data and evaluation metrics.
- Strong spoken English - B2+ or higher for discussing results and trade-offs with international teams.
Nice to Have
- Experience with YOLO, Detectron2, MMDetection, Segment Anything, or similar tools.
- Experience with ONNX, TensorRT, CoreML, or edge inference optimisation.
- Experience with retail analytics, manufacturing inspection, OCR, video analytics, or camera-based products.
- Experience with MLflow, Weights & Biases, DVC, or dataset versioning tools.
Apply
If you enjoy applying computer vision to real-world visual data with practical deployment constraints, we would be glad to hear from you. Send us your CV and we will contact you to discuss relevant opportunities.
Our success stories
Tell us about your project needs
.png)
.png)
