Quadrature

Quadrature is a systematic trading company, in the business of building fully automated trading systems across the global financial markets since its founding in 2010. We're a very tech and research focused org, and are quite different from a lot of other players in the market (more like a tech-driven research institute than a traditional trading firm -- see https://quadrature.ai for the details). We're looking for a skilled engineer to optimise the use of GPUs for our AI workloads, for both our model training and inference environments. This is a great opportunity to work with a group of world-class researchers and developers, building out large-scale algorithmic trading systems using cutting-edge tech (advanced ML, thousands of GPUs, multiple global research data centres and trading co-location sites, petabytes of storage, innovation-lab R&D work with strategic partners). It'll be based out of either London or New York. Responsibilities: - Leading performance optimisation effort for mission-critical end-to-end AI workloads (e.g. profiling and tuning GPU programs, writing custom CUDA kernels, speeding up pre- and post-processing). - Rewriting core parts of our model architectures for maximum performance. - Building a runtime environment that maximises the model capacity and speed. To learn more about what we can offer have a look at https://quadrature.ai/careers/benefits Feel free to email me directly via labeed@quadrature.ai