Jeff Farris

Jeff Farris is a principal researcher at NVIDIA on the Nemotron data, scaling and evaluation team. He develops evaluation methods that use early training signals to predict model capabilities and guide data and training decisions. He co-developed SWE-Serve, a benchmark that evaluates AI coding agents on real inference-serving tasks. Before NVIDIA, he led automatic speech recognition model development and the deployment of AI systems at scale at AWS and for the U.S. government.