Nicholas Egan is an Applied Research Scientist at Primer AI, where he teaches machines to read and write. He is responsible for creating Primer's pronoun coreference resolution model, Bayesian and bandit based hyperparameter optimization tools, the open-source reference implementation of BLANC (Primer's human-free summarization evaluation method), models that find factual dispute and infer agreement or contradiction, and Primer's methodology for explaining decisions of transformer models. Before coming to Primer, Nicholas worked on GANs at MIT's Computer Vision Lab, and interned at Robinhood, Facebook, and Airbnb.
