ABBYY has launched a new self-service application programming interface (API) designed to help developers extract accurate data from business documents with minimal coding effort.

Known as ABBYY Document AI, the API addresses growing challenges faced by developers who need to transform unstructured business documents into structured, reliable data. According to ABBYY, the solution allows users to integrate powerful optical character recognition (OCR) and intelligent document processing (IDP) capabilities using just a few lines of code.

"As a vanguard of OCR, ABBYY has long had a vibrant community of cutting-edge developers creating transformational solutions with our advanced document AI," Nick Hyatt, Vice President of Engineering R&D at ABBYY.

"We are providing them a new API with minimal setup, access to ample community resources, and pre-trained models for building proof-of-concepts."

The intelligent document processing market is experiencing significant growth, with IDC projecting an expansion from $US2.4 billion in 2023 to $US10.5 billion by 2028, representing a 34.9% compound annual growth rate. This growth is attributed to increasing cloud adoption, AI maturation, and expanded document AI use cases.

Amy Machado, Senior Research Manager at IDC, noted that OCR is experiencing a "true renaissance" in the age of AI. However, developers using general large language models for document processing often encounter challenges such as hallucinations, data inconsistencies, and errors, particularly when dealing with multiple languages, handwriting recognition, and complex document structures.

The ABBYY Document AI API, initially available as a technical preview, offers pre-trained models to extract data from documents and accelerate automation for complex business processes including KYC (Know Your Customer), account openings, customs clearance, invoice processing, expense management, and order processing.

A key feature of the new API is its ability to preserve a document's logical structure, providing AI-ready data that can be used for generating insights in generative AI applications and retrieval augmented generation (RAG) systems, or for training language models.

The company has made comprehensive software development kits (SDKs) available for Python, C#, JavaScript, and Java. Developers interested in early access can join the preview list through ABBYY's website.