TL;DR We ran on the DocVQA validation split and got 5,286 correct out of 5,331 (99.15%). Of those 45 wrong answers, only 18 are true parsing shortcomings. DocVQA is usually used to evaluate vision-language models, but we are pioneering the use of this popular dataset to establish the accuracy of our Agentic Document Extraction (ADE) […]
Ava Xia
November 7, 2025
Signatures, stamps, and seals are the final proof of authenticity in documents, but they are also among the hardest elements to detect automatically. With the new attestation chunk in Agentic Document Extraction (ADE), these visual markers are identified, structured, and extracted with spatial context intact. From healthcare reports and financial forms to legal affidavits, ADE ensures every attestation is captured accurately and ready for reliable, audit-ready automation.
Qiang He
November 6, 2025
In the era of large language models, a silent barrier has emerged for non-English speakers. Despite the revolutionary capabilities of modern AI, there’s an uncomfortable truth: most mainstream models are trained predominantly on English corpora, creating an implicit performance gap that affects text parsing, information extraction, and document processing across languages. For organizations processing Chinese […]
Nearly every enterprise runs on documents. Legal agreements, claims packets, forms, loan applications, contracts, just to name a few. These documents can be long, messy, complex, and multimodal, including scanned forms, nested/complex tables, charts, graphs, engineering & architecture diagrams, checkboxes, radio buttons, images, signatures, footnotes and much more. Often, business users spend hours re-keying or […]
Bianca Ragsdale
October 28, 2025
Your team processes hundreds of property insurance claims daily. Some arrive as PDFs, others as scanned images. One claim might list the policyholder's name as "John Smith," while another uses "Smith, John." The data you need is in there somewhere, but getting it out consistently? That's the challenge.
Dan Maloney
October 28, 2025
40% of Loans Are Referred to Manual Review—Often 500+ Pages. Automated Document Intelligence Unlocks the Workflow. In this blog, we’re exposing the multi-billion-dollar bottleneck in the mortgage industry. It’s not about interest rates; it’s about the massive, hidden pile of “complex” loans that your automation can’t handle. These are the loans that get kicked out […]
Andrea Kropp
October 17, 2025
Identity documents like passports, driver’s licenses, and green cards are difficult to automate because of complex layouts, security features, and compliance risks. Traditional OCR and templates fall short. ADE provides accurate, auditable, and scalable extraction with schema-driven, layout-aware parsing and visual grounding.
Ava Xia
October 16, 2025
Imagine your application needs to process data from a massive 1000-page contract or a 1-GB engineering report scanned into a single PDF. The upload completes, but now the parsing step grinds everything to a halt, blocking your app, slowing users down, and backing up your processing queue. Developers working with real-world documents know this pain […]
DeepLearning.AI
October 11, 2025
The Batch AI News and Insights: LandingAI’s Agentic Document Extraction (ADE) turns PDF files into LLM-ready markdown text.
Join our Newsletter
Subscribe to our Newsletter to receive exclusive offers, latest news and updates.
