ABBYY unveils AI tools to enhance document processing

ABBYY has announced the release of several new artificial intelligence (AI) capabilities designed to enhance business operations related to document-specific processes.

The newly introduced tools are included in ABBYY's Purpose-Built AI Center, which serves as a hub for businesses and developers looking to use AI to develop next-generation applications. Among the key features is 'Phoenix', a multi-modal approach to zero-shot learning that leverages small language models specifically designed for document tasks.

Other noteworthy capabilities include a Secure Large Language Model (LLM) Gateway and a collection of enterprise-ready tools and applications that provide global enterprises with enhanced insights into business-critical document processes.

The Purpose-Built AI Center is intended to simplify the onboarding of new document types into enterprise business processes, offering businesses greater flexibility to adapt to changing needs. This centre acts as an information resource detailing ABBYY's array of AI tools, including insights into how small language models can deliver the capabilities of larger language models with increased speed to value.

ABBYY utilises intelligent document processing as a secure LLM gateway. This allows businesses to extract validated data from documents while minimising the risk of hallucinations, thereby increasing the trustworthiness of the output.

Among the new tools, ABBYY has introduced several capabilities to optimise operations for large enterprises. One key feature is e-invoice processing, which offers a streamlined and compliant method for handling various invoice formats according to diverse national standards. Additionally, ABBYY has developed quality analytics tools specifically designed to analyse trends related to straight-through processing (STP) rates, facilitating continuous improvement in data extraction quality and performance.

Furthermore, the company has enhanced its big data analysis capabilities, providing improved scalability for process analysis and enabling the management of complex, event-driven data involving billions of data points.

Maxime Vermeir, Senior Director of AI Strategy at ABBYY, commented, "With the launch of our Purpose-Built AI Center, we are making it easier for our customers to find the right solution for their business needs and simplify their success in automation. ABBYY's innovations truly respond to the main concerns businesses have in the market today when leveraging AI and doing so at scale within an enterprise context."

In addition to the enterprise tools, ABBYY also introduced advancements in optical character recognition (OCR) performance aimed at aiding developers in the creation of advanced applications. The enhancements include improved memory consumption for OCR, which boosts document conversion quality and structure detection. They also extended handwriting recognition across multiple languages, introduced new document comparison modules, and provided updated processing profiles for data extraction, natural language processing (NLP), and archiving.

Support for APIs has been streamlined, with new and updated support for widely used development languages such as .Net and Python, among others.