LandingAI, a leading innovator in agentic vision AI technologies, announced a major upgrade to its Agentic Document Extraction (ADE) platform with the introduction of Document Pre-trained Transformer-2 (DPT-2). The new model is designed to extract information with unparalleled accuracy, even from complex documents, empowering organizations to make faster, data-driven decisions.
Building on the success of ADE’s initial launch six months ago—which enabled developers, startups, and Fortune 500 companies to process billions of pages and reported up to a 90% reduction in time spent searching for information—DPT-2 sets a new benchmark for document intelligence.
Unlike generic large language models (LLMs) that often struggle to fully and consistently extract visual and structured data, DPT-2 combines deep learning with agentic workflows to deliver highly accurate and trustworthy outputs. It can process documents containing tables without gridlines, invoices scanned at unconventional angles, embedded signatures, check marks, and other challenging formats.
Also Read: Sycurio Appoints Nicholas Bown as Chief Security Officer
“Documents contain the information that organizations need to make not only accurate but the best decisions possible. Key nuances can be lost if the visual representations are not adequately captured,” said Dan Maloney, CEO of LandingAI. “ADE addresses this gap and our enhanced release pushes capabilities even further.”
LandingAI anticipates strong adoption in industries such as finance, healthcare, insurance, and compliance, where accuracy is critical and workflows are ready for enhanced agentic efficiencies.
“We’re past the era of one-size-fits-all models. Just as we use different processors today for different workloads (your smartwatch has a less powerful processor than your laptop than the datacenter), for AI workloads too, we need different types and amounts of intelligence to do different things,” said Andrew Ng, founder of LandingAI. “ADE DPT-2, specialized in processing documents, works uniquely well in the space of document extraction.”
The ADE DPT-2 release introduces several new features to simplify complex document parsing:
-
Agentic Table Captioning: Accurately parses large, no-gridline, and merged-cell tables while preserving cell-level details, enabling users to trace values back to their source.
-
Refined Figure Captioning: Precisely identifies logos, seals, and small figures while eliminating excessive descriptions.
-
Smarter Layout Detection: Improves recognition of all document sections, including stamps within tables, crucial for compliance workflows.
-
Expanded Chunk Ontology: Classifies signatures, checkboxes, ID cards, barcodes, and QR codes consistently alongside text, tables, and figures for comprehensive document understanding.
With ADE DPT-2, LandingAI continues to redefine the possibilities of AI-powered document intelligence, enabling enterprises to harness complex data with greater speed, accuracy, and confidence.
