CereSoft was founded on idea that you can turn documents into data with intelligent technology. In 1990, a group of professors at the University of Maryland envisioned an intelligent software system that could extract information from any document image and transform it into editable electronic data. They formed a company called Nonlinear Technologies Inc. (NTI) and developed a handwriting recognition system. The United States Defense Advanced Research Projects Agency (DARPA) awarded NTI a research grant to turn their system into a commercial product. 

This project was a resounding success and attracted the attention of the established software company OCR Systems, Inc. OCR Systems teamed up with NTI to produce an OCR engine for the recognition system. In 1992, this first generation recognition engine was entered into a benchmark performance study conducted by the United States National Institutes of Standards and Technology (NIST) and the National Census Bureau (NCB). NIST tested all of the world's available handwriting recognition systems and found NTI's engine to be the most accurate. 

This notable distinction brought NTI to the attention of Adobe Systems, Inc., who wanted to use the recognition engine to support a new software product called Adobe Acrobat. NTI's staff became part of the development team for the Adobe Acrobat Capture product. 

After releasing the Adobe product, NTI reorganized into CereSoft, Inc. in 1994. CereSoft introduced a system integrated with high level linguistic models to deliver the best possible data capture results. CereSoft also developed a revolutionary Universal Script that uses language and document cues to capture data from any document regardless of format.