BriefOur inference systems need to act fast, taking 1.4 million decisions a day with incredibly low error rates and running on nothing more that the edge resources available. As lead engineer for AI you will develop and implement novel material recognition solutions using the data deciphered from multi-energy X-ray during high speed industrial inspection, where even small anomalies can present a high safety risk.You will want to join a small, multidisciplinary science and engineering team dedicated to commercialising cutting edge X-ray inspection technologies with genuine market benefits.You will have a passion for hands-on projects and building scalable solutions.You will be curious about how things work and how to maximise their performance.You will enjoy both autonomous and collaborative working alongside business leadership and PhD colleagues in microelectronics and high energy physics in a dynamic technology spin-out.ResponsibilitiesLead the deployment and optimization of edge-based neural networks for real-time inference on resource-constrained devices.Design and implement scalable deployment pipelines, including model conversion, quantization, and validation for GPU-based edge platforms.Monitor performance metrics and conduct rigorous testing to ensure reliability, low latency, and power efficiency in deployed models.Collaborate with cross-functional teams to evaluate new hardware platforms and deployment tools for future model compatibility and performance gains.Maintain documentation and best practices for deployment processes, toolchains, and model optimization techniques to support long-term scalability.Skills - EssentialBachelor's or Master’s degree in Computer Science, Electrical Engineering, AI Engineering, or a related technical field (PhD is a plus).Strong experience deploying deep learning models in production environments, with an emphasis on high-throughput, low-latency inference.Hands-on experience with modern ML frameworks such as PyTorch and Keras.Strong proficiency in modern C++.Experience with the C++ API of PyTorch (LibTorch) for deploying models in production or embedded environments.Proven expertise in optimizing models for GPU inference, including:TensorRT, ONNX Runtime, TorchScript, or TVMMixed-precision inference (e.g., FP16, INT8 quantization)Batch optimization and asynchronous executionStrong collaboration skills to partner with AI researchers in bridging research and realworld embedded applications.Skills - BeneficialKnowledge of high energy physics and X-ray behavioursKnowledge of image reconstruction from multi-level and multi-projection data setsGood communication, both as a collaborative team player and critical thinkerGood problem solver, able to anticipate issues and create solutionsStrong project management, presentation and report writing skills (professional English)Start Date: from June 2025Location: Cerdanyola del VallèsRole Type: Full TimeLanguages: English (required), Spanish