Oosto (formerly AnyVision) is a world leading team of scientists and engineers that creates AI for the Real World. With deployments and clients spanning the globe, we are recognized as leaders in the fields of Face Recognition, Object Recognition, Object Detection, and Visual Target Tracking.
Oosto is dedicated to providing a diverse and flexible work environment so that we can recruit and retain the best talent.
The research optimization team is building cutting edge AI platform to deliver the most efficient inference to wide spectrum of devices at scale in production. Our vision is to offer end-end fully automated solution to developers and researchers, with which they will be able to optimize and deploy AI capabilities to any hardware.
The role will offer broad opportunities to be exposed to various AI systems and hardware, and you will enjoy working on with cutting-edge optimization research and development that can be implemented and deployed to our AI products.
You are:
You are a highly motivated individual with a rigorous, meticulous, and scientific approach to life. You have completed a degree in a scientific discipline, with some emphasis on programming. You should be comfortable working within a team when it suites the project at hand, but also be capable of working on your own if needs be. You are creative and passionate about large-scale research problems with very professional engineering experience to deliver your work to the field.
You will:
l Work on a platform for ML model optimization and deployment.
l Design, build and manage processes to deploy production models and monitor their quality and performance.
l Focus on making deployment manageable and scalable from edge device to cloud.
l Develop automation strategies for deployment of ML models.
l Research on machine-learning-based optimizations for better performance and efficiency.
l Explore and develop hardware-aware optimization algorithms in our platform.
l Improve the team and yourself through brainstorm, design review, code review, etc.
Requirements:
Must have skills:
l Experience implementing or optimizing linear algebra operators.
l Strong Python and C++ skills (highly template programming).
l Experienced on deep learning frameworks e.g., Caffe, PyTorch
l Knowledge or an awareness of GPGPU programming models like CUDA, OpenCL, SYCL
l Familiar with computer architecture, low-level optimisation, via compilers or manual.
l Experienced in ML model optimization techniques (compression, quantization, etc).
l Familiar with software development lifecycle.
Beneficial Skills
l Developing experience on inference engines e.g. OpenVino, TensorRT.
l Familiar with deep learning compiler stacks such as TVM, TC, nGraph.
l Working knowledge of compilers and compiling techniques (Clang/LLVM)
l Android/IOS development experience.
UK, Belfast, London,
