Route Reports sells end-to-end solutions for transport and infrastructure. In order to provide the best solutions to our customers, our hardware and software is all developed in-house. Our customers include the largest transport and infrastructure companies in the US and UK. Our solutions focus on giving transport providers and infrastructure owners the information they need to maintain their infrastructure or run their services more efficiently through the use of real-time analytics, computer vision, and predictive insights. Our products are usually among the first to market and are quickly gaining significant traction with major customers.
Our UK offices are at Southbank in Central London. We offer a modern office with several amenities nearby and flexible working options.
We are looking to enhance the existing computer vision algorithms for our current products; all of which combine machine learning with vehicle-mounted cameras. We are looking for a proactive candidate with a strong understanding of machine learning concepts and an enthusiasm for applying their work to real life situations. To be successful you should be able to show off evidence of your personal deep learning projects. You could be responsible for developing & training neural nets, optimising for embedded devices, testing or deploying the system, and ensuring that the results generated match customer expectations.
- Research, design, and develop computer vision algorithms for embedded devices.
- Working alongside our Head of Product to develop new computer vision algorithms based on customer requirements.
- Working alongside data engineers and data labellers to improve training data sets.
- Interacting with other members of the tech team to ensure a complete system design.
- Optimising computer vision solutions for embedded hardware.
- High skill level in Python and/or C/C++.
- Prior experience in developing computer vision solutions;
- Proficiency with machine vision and machine learning frameworks (OpenCV, TensorFlow + Keras, TensorRT, PyTorch, Pytorch + FastAI etc) with a strong portfolio of development examples;
- Understanding and prior application of basic Machine Learning principles (Precision Recall Tradeoff, Accuracy metrics, Under vs Overfitting, Clustering, SVMs, etc)
- Comfortable reading, discussing, and applying research from published papers
- Experience training Neural Net architectures for classification, object detection, and segmentation
- Inference deployment (Preferred)
- Experience with FP-32, FP-16, or INT-8 training and inference (preferred)
- Experience with cloud-based training and deployment pipelines (preferred)
- A pro-active, self-managing attitude