The company’s new Visual Document Understanding (VDU) language model marks a significant advancement in enterprise data extraction and classification.
New York, NY – November 1, 2023 – Hyperscience, a provider of enterprise artificial intelligence solutions, today announced that it has been awarded a 2023 Innovation Award from Deep Analysis, a respected analyst firm in the unstructured data automation field. Commended for its research on generative document AI models, Hyperscience was singled out for its ability to eliminate the necessity for Optical Character Recognition (OCR) while still maintaining remarkable document comprehension.
“Hyperscience has successfully transitioned from a cool VC-funded AI start-up into a mature enterprise AI company, working with top partners and winning big deals with blue chip customers,” the Deep Analysis Innovation profile detailed. “With its latest release, Hyperscience has a highly advanced, end-to-end IDP platform for all document classes. The company developed its own OCR engine from the ground up, eschewing traditional engines developed in the 1980’s and 1990’s that are used in Kofax, ABBYY, and other classic IDP vendors. After 50 years of OCR engines ruling the document processing roost, Hyperscience is an innovation leader for an exciting new data extraction and classification phase.”
Historically, traditional Intelligent Document Processing (IDP) solutions rely on OCR to transcribe, classify, and extract documents, requiring multiple machine learning models to process a document end-to-end. This multi-model process, while being pivotal for some use cases, also presents several drawbacks, including compounding model errors, model management overhead, and slower processing times.
The Hyperscience engineering team experimented with a new approach, using a generative document AI model that reduces complexity and accelerates time to value for customers. The team used an OCR-free visual document understanding model to dramatically reduce the number of document processing steps, streamlining end-to-end document processing. Now, customers can integrate machine learning models more seamlessly into their automation processes, minimizing text errors, and expediting the time it takes to go from the beginning of a project to its production phase.
“Generative AI is a disruptive technology that will transform how organizations leverage data to automate processes, increase productivity, and strengthen customer engagement,” said Andrew Joiner, CEO of Hyperscience. “Hyperscience stands out in the market with a modern, AI-based approach for processing data, our ability to automate enterprise business processes securely and at scale, and our steadfast focus on driving meaningful customer outcomes. We are honored to be recognized with the Deep Analysis Innovation Award, spotlighting our commitment to driving end-to-end hyperautomation.”
To learn more about how Hyperscience uses generative document AI models without OCR, see the Hyperscience blog: Using Out-Of-The-Box Generative Document AI Models.
At Hyperscience, we believe that the dawn of AI represents tremendous potential for enterprises to transform their operations and unlock new levels of efficiency and insight. Our mission is to bring the opportunity of AI to the enterprise, empowering our customers to automate their most complex, mission-critical processes with ease. We’re committed to delivering solutions that not only meet the highest standards of performance, but also inspire trust and confidence in our customers.
Hyperscience is relied upon by the world’s most respected organizations such as Mars, The United States Department of Veterans Affairs, Guardian Life, The International Rescue Committee, and HM Revenue and Customs. Hyperscience is backed by leading investors including Bessemer Venture Partners, Stripes, FirstMark, Battery, and Tiger Global.