Article: Cannata Report - April 12, 2024
As organizations continue to chew and choke on the challenges of managing extensive data volumes, artificial intelligence (AI) is becoming pivotal in shaping the future of document management systems (DMS).
Accessibility to text-synthesis machines is still new. Chatbots and generative AI burst into mainstream consciousness 18 months ago when ChatGPT took the internet world by storm with its ability to answer questions and generate content in a human-like way. Quickly following to market in early ’23 was Google (Bard, now Gemini 1.5) and Microsoft (Copilot, the consumer’s “everyday AI companion,” per last month’s Super Bowl commercial for the MS app).
Financially speaking, 2024 is predicted to be a telling year for these and other big-data powerhouses, including Apple Inc. and Meta Platforms, as well as AI chipmakers Advanced Micro Devices, Inc. (AMD) and NVIDIA Corp. All six publicly held companies are demonstrating optimism by investing heavy R&D dollars into advancing gen-AI technology.
AI, while admittedly disruptive, is not truly smart and is far from emergent. IBM debuted its Watson natural-language system 14 years ago. But “thinking machines” and large language models (LLMs) have been in use since the 1950s, according to experts at Harvard University. LLMs are deep-learning models pre-trained on vast amounts of data. Alex Wong, senior product marketing manager at Laserfiche, points out that his firm, which specializes in enterprise content management (ECM), has been involved with predictive AI for quite some time. “Smart invoicing, OCR, and voice-to-text transcriptions are nothing new,” he said.
GEN AI APPLICATIONS
Generative AI applications, however, are newer. This type of artificial intelligence is capable of generating text, images, or other data using generative models, often in response to prompts. Laserfiche dipped a toe into these waters during Q4 2023. Built into the firm’s cloud-computing platform, the RegEx Builder tool uses gen AI to help users natively build and update regular expressions with much less coding.
Rather than build solutions from scratch, the new Laserfiche product can be used to accelerate the process of turning ideas into usable pieces of code that, in turn, can be integrated into forms. The process substantially shortens the time from idea to delivery. “Users just type in regular expressions, and the Builder will proofread,” explained Wong. “It can validate form fields for categories like phone numbers, ZIP codes, or U.K. postal codes.”
NEXT-GEN DOCUMENT PROCESSING
“As classic document management systems morph into intelligent document processing (IDP) solutions, we are seeing the addition of several advanced technologies, not the least of which is AI,” said Anne Valaitis, a principal analyst for Keypoint Intelligence’s AI Advisory Service. “Artificial intelligence is especially useful in the capture modules many DMS products offer to help knowledge workers get documents into the system.”
She explained, “With AI, documents can be classified by the system automatically—based on content, contextual clues, and other learned information—instead of based on reading predetermined metadata fields or OCR parsing alone. Once data and documents are in the platform, AI can assist in any document conversion or processing that is desired, as well as in automating complex workflows.”