james k reed
Systems Software Engineer with a Focus on Deep Learning Systems
James K. Reed is an experienced software engineer with hands-on experience in programming languages, compiler design, and deep learning platforms. He is fascinated by the whole computing stack from programming interface down to silicon, and seeks to investigate the intersection of these as the newest AI revolution drives High-Performance Computing (HPC) innovation.
At his current role at fireworks.ai, James is applying his skills to build the fastest, most extensible, and easiest-to-use platform for multimedia inference, including image, video, and audio signal modalities. Before this, James worked with the PyTorch team at Meta (formerly Facebook) as a Staff Software Engineer, where he led development of toolkits including torch.fx, PiPPy, TorchScript, and ONNX. He completed his Bachelor of Science in Computer Engineering from Virginia Tech, graduating Summa Cum Laude.
In his free time, James has taken to learning languages, with a particular focus on Spanish and Italian. He is also fascinated by the deeper linguistic aspects of natural language, including etymology, phonology, grammar, and their evolution over time. Parlami, amico mio!
You can follow James’s work on his GitHub profile and connect with him on LinkedIn. James has published academic papers, patents and blog posts and has given several talks.