Founded by Dr Peter Baldwin in 2017, Myrtle is a software company that takes machine learning models and converts them into specialised software for deployment on field programmable gate arrays (FPGA). FPGAs are often much faster and more efficient than GPUs for various tasks, but converting and optimising machine learning models to run on them is complex and time consuming. Myrtle enables their customers to take advantage of the power of FPGAs without the need for this capability in-house. This can alleviate the compute bottleneck of AI inference at scale, and unlock the application of machine learning models in ultra low latency settings.