Back to the newsroom

Press Release

Korea Deep Learning unveils and demonstrates VLM-based DEEP OCR+ at World IT Show 2025

Korea Deep LearningPublished

Korea Deep Learning unveils and demonstrates VLM-based DEEP OCR+ at World IT Show 2025
In April 2025, Korea Deep Learning unveils VLM-based DEEP OCR+ at World IT Show 2025 with a demonstration that processes documents on the spot.

Korea Deep Learning will take part in World IT Show (WIS) 2025, Korea's largest ICT exhibition, held at COEX in Seoul from April 24 to 26, 2025, and unveil DEEP OCR+, a document AI solution built on a hyperscale vision-language model (VLM). Centered on DEEP OCR+, which it launched on April 10, 2025, Korea Deep Learning will demonstrate on site how a VLM reads and structures unstructured documents. The plan was announced on April 23, 2025.

DEEP OCR+ is a document AI that grasps the structure and meaning of a document to extract and summarize key information. It recognizes many document formats immediately without separate labeling or templates, and exports results as JSON or Excel so they can be connected to ERP and RPA systems. For example, when a contract is uploaded, it automatically extracts the contract date, the amount and risky clauses. It can be adopted as SaaS, through an API or on premises, and can be put to work in the field within two weeks on average. By processing documents that visitors bring on the spot, the demonstration shows firsthand how it reads documents in all kinds of formats without prior training.

Conventional OCR focused on extracting text and was limited in handling complex structures such as tables, clauses and paragraphs. Contracts and invoices in varying formats required new design and labeling for each document, driving up the cost of adoption and maintenance. In a February 2025 interview with ZDNet Korea, Korea Deep Learning said that the role of OCR grows as document digitization spreads, and that it would make OCR the core of business automation rather than a document recognition technology. DEEP OCR+ is the result of turning that direction into a product.

DEEP OCR+ is built on a VLM trained on more than 400 million text and image documents. It processes many types of documents without separate data collection or labeling, and once a document is uploaded, it returns the key information in structured form with no additional training. When it launched DEEP OCR+, Korea Deep Learning named finance, law and the public sector as the main target areas. Korea Deep Learning said that DEEP OCR+ is being considered for adoption in a range of fields, including finance, law and the public sector, and is already in pilot use in some large projects.

The show is the first major exhibition to follow the launch of DEEP OCR+, and a chance to gauge the market's response. According to Korea Deep Learning, as of April 23, 2025, average monthly inquiries over the previous two months had risen more than 15-fold.

Korea Deep Learning counts more than 80 public and private organizations as customers and has secured 11 partner companies in Japan. It holds 13 AI patents along with TTA and ISO 9001 certifications.

About DEEP OCR

Korea Deep Learning started out by reading documents. Built on KDL's own document-specialized VLM, DEEP OCR is Korea's first VLM-based OCR solution, reading not just the characters but the structure and meaning of a document together. It handles unstructured documents in every kind of layout without building separate training data, and extracts text and key-value data from visual elements such as seals, signatures, tables, numbers and symbols. From loan documents in finance to administrative documents in the public sector and handwritten waybills in logistics, documents people used to copy out by hand become business data through DEEP OCR. More information is available on the DEEP OCR page at koreadeep.com.

About Korea Deep Learning

Korea Deep Learning (KDL), founded in 2019, is an enterprise AI agent company. Built on KDL Frontier, its own vision-language model (VLM) specialized for documents, KDL makes AI agents that take on and complete the document work of businesses and public institutions. Its main products are DEEP OCR, DEEP Parser, DEEP Agent and DEEP Ops.

Source