Anthony • Dec 14, 2021

A startup company I co-founded in 2016 while I was consulting between jobs. We explored using natural language processing techniques on US patent grants and Pubmed articles and built a couple prototype products around this technology. We made heavy use of Explicit Semantic Analysis and Locality-Sensitive Hashing techniques to do high-throughput, low-latency query-to-document matching on midsized corpora (~100 million documents), with the intention of enabling real-time exploration via computer-aided authoring. I separated from the company in late 2019 and it eventually shuttered in 2020 or so; only an old Internet archive snapshot remains of the company's web presence. I retained ownership of the some of the IP I produced for the company, described in Legit techLegit tech
The primary goal of this set of modules is taking text processing capabilities up one level of abstraction. Many of the items listed below already exist as freely-available libraries. These librari...