JOSEPH
DAVISON

WHAT DO YOU DO?

I’ve been immersed in AI since my early coding days, pioneering neural network architecture search during my undergrad at BYU. At Microsoft, I extended their deep learning framework to R users. My master’s research at Harvard focused on “prompting,” now a ubiquitous technique. In NYC, I played a key role in making Hugging Face the “GitHub of Machine Learning.” Recognizing the potential of large pre-trained models, validated with GPT-3 in 2020, I then joined Enveda Biosciences in Utah. Now, as a PhD student at the University of Utah and with my startup, Neural Assembly, I’m driving the next era of computing abstractions for LLM.

WHAT DOES YOUR COMPANY DO?

At Neural Assembly, we’re crafting the Microsoft of the AI era, transforming models into intelligence. LLMs are processors — we’re developing the OS.