In this scenario, pioneers of next-gen AI, Ecompex is stirring a revolution in information management through their full-lifecycle of document digitization and content management services for the government, based on AI solutions that can learn, evolve, act, and perform, much like how the human brain works. As opposed to an algorithm-based system that is specific and follows a data trick, Ecompex uses an inference-based engine that is configurable according to a knowledge base and improves by learning. “Moreover, while an algorithm-based system falls short in terms of flexibility and scalability in the face of large-scale decision-making processes, our AI applications are the perfect fit when it comes to enhancing the integrity, effectiveness, and economics of large-scale governmental administrative processes that need quick, accurate, and consistent decision making,” says Hubin Jiang, President and CEO of Ecompex.
Founded with the sole mission to spearhead digital transformation for governments, Ecompex’s AI journey is truly inspiring. Initially delivering AI applications for document classification and extraction, the visionary thinkers at Ecompex advanced into document comprehension and information and intelligence extraction in 2010. Today, the company develops applications for document understanding that tells a story and displays human-like tacit knowledge rather than a system that is programmed to simply read a story.
To inject further light into Ecompex’s value proposition, let’s assume a scenario. A records manager with access to millions of enterprise records is faced with the herculean task of classifying the documents into a taxonomy, which contains thousands of different classification codes. This is to be followed by the extraction of structured data from the classified documents, comprehension of what the documents are talking about, and then processing of information according to specific project needs. To add to the challenge, the documents could be anything from medical to legal or administrative documents; as well as in any format such as forms labeled with organizational codes, a partial format such as a letter or an email, electronically-generated documents, or poor-quality scans of hard-copy documents.
How can Ecompex’s state-of-the-art intelligence technology help? The Ecompex system allows records managers to describe the documents, whether they are medical records, legal documents, or financial records.
Behind Ecompex’s ability to deliver high-level understanding from textual content is their futuristic AI-based content comprehension software, IntelliDigo. IntelliDigo delivers rapidly customizable, learning-based text extraction to provide a deeper understanding of content that can be pulled from multiple sources of unstructured text. IntelliDigo is a text exploitation platform that is comprehensive, flexible, scalable, and web-based for multisource text analysis and can work with open source frameworks such as Apache Unstructured Information Management Architecture (UIMA).
The company develops applications for document understanding that tells a story and displays human-like tacit knowledge rather than a system that is programmed to simply read a story
Today, a pertinent challenge in each expertise domain is that to achieve maximum results, traditional products must rely on customizations for specific use cases. This ties up Subject Matter Experts (SMEs) as they have to spend valuable time understanding the underlying technology in order to perform knowledge capture. IntelliDigo, on the other hand, uses inference engine technology that allows domain experts to quickly, affordably, and effectively transfer knowledge to a machine without any system customizations. This results in SMEs being able to concentrate on what they do best, which is knowledge capture and data analysis. In addition, IntelliDigo is developed to extract and associate event arguments with events, extract relations between events, and understand how events fit together in a logical way.
To that end, it’s evident why Ecompex has been the solution-of-choice for the U.S. Department of Justice (DOJ) for large-scale and high visibility litigation support projects. In support of DOJ, Ecompex provides imaging, coding, and proofing of over 10 million pages per year and for 9,000 different types of documents at a turnaround time of 24-72 hours or even less. “Although compliance requirements such as NARA, NIST information security, HIPPA, and DOJ compliance pose challenges, we ensure high accuracy for all classification and coding. Moreover, it is impossible to manually remember 9,000 variants of documents—when they are mixed—and code consistently and accurately,” remarks Jiang. Ecompex uses its AI technology and application, IntelliDigo, to pre-classify and pre-analyze documents and images, automatically identify document type and data to be coded, and guide coders to pick up the data and information accurately.
In the near future, Ecompex is planning to expand into the private sector as well as core domains such as telemedicine, social networking, and identity management. On the product side, they will be developing their APIs to offer wider capabilities by integrating more applications. “The future has tremendous prospects in store, and the public sector is a huge market to explore. We will expand Ecompex’s robust technology into every corner easily,” ends Jiang.