Press Release
Korea Deep Learning closes KRW 12 billion Series A with KRW 5 billion from Korea Development Bank
Korea Deep LearningPublished

On January 8, 2026, Korea Deep Learning closed its Series A, totaling KRW 12 billion. Korea Development Bank closed the round as the final investor, putting in KRW 5 billion, which accounts for more than 40% of the total. Moving toward becoming an enterprise AI agent company, Korea Deep Learning will focus the funds on advancing document understanding AI, workflow automation and global expansion.
A round that passed its KRW 10 billion target
The round was led by Translink Investment, with SV Investment joining as a major co-investor. Translink Investment, SV Investment, IBK Industrial Bank of Korea and Korea Development Bank took part in the first close in October 2025, and Korea Development Bank then raised its planned investment from KRW 3 billion to KRW 5 billion, pushing the final amount raised past the KRW 10 billion target.
Since its founding, Korea Deep Learning has grown on its own revenue alone, with no outside investment apart from a KRW 100 million TIPS grant. This Series A is its first large-scale funding since it was founded. That its first investment came to KRW 12 billion shows how it has grown until now.
What the investors saw: the ability to commercialize
Korea Development Bank rated highly its real ability to commercialize, including more than 80 customers, KRW 10 billion in cumulative revenue, and a productization strategy that lets customers deploy within two weeks without separate training. Translink Investment, the lead investor, described Korea Deep Learning as a team in the AI field with Vision-LLM technology and the ability to execute quickly.
The investment industry sees unstructured data as the bottleneck in enterprise AI adoption. Companies want to build custom AI with their internal data, but most of their documents are unstructured data mixing images and tables, which makes them hard to process. SV Investment explained that a bottleneck is forming in the process of training models on data, and that Korea Deep Learning has tackled this problem head-on. Korea Deep Learning was recognized for bringing multimodal technology, such as vision-language models (VLM) that understand text and images together, to product faster than anyone else in Korea.
Korea Deep Learning has around 80 customers, including public institutions and large companies, and is also a member of Inception, NVIDIA's startup program.
KRW 1.1 billion in contracts two months after launch: the results that drew the investment
Behind the investment is DEEP Agent, the document-based AI agent launched in March 2025. DEEP Agent won contracts worth about KRW 1.1 billion within two months of launch, and orders in the first half of 2025 grew 397% from the same period a year earlier. Customers' deployment costs fell from an initial KRW 300 million to KRW 150 million, and build time from 66 days to 10 days. Following the passing of KRW 10 billion in cumulative revenue in December 2024, outside investors confirmed a pattern of the product turning into order growth.
From document automation to workflow automation
Where the funds will go is also clear. Korea Deep Learning will concentrate the funds on advancing its document understanding AI technology, building systems that automate the work that follows, strengthening AI execution rules and verification, optimizing processing speed and cost, and entering global markets. It will also use them for core technology R&D and recruiting top talent.
In 2026, it will expand its business beyond document automation into "workflow automation." In the first half, it plans to expand industry pilots, and in the second half to launch a full AI agent product. The goal is for AI not only to pull information out of documents but to handle the underwriting, verification and entry work that follows from that information.
"Closing this Series A is an important turning point for Korea Deep Learning as it moves beyond 'AI that reads documents' to 'AI that finishes the work with documents.'"
— Kim Ji-hyun, CEO, Korea Deep Learning
Founded at 22, grown on its own revenue for five years
Korea Deep Learning was founded in 2019 by CEO Kim Ji-hyun at the age of 22. For close to five years after its founding, it built up revenue without outside investment and passed KRW 10 billion in cumulative revenue. Its flagship product is DEEP OCR, which grasps even the context of documents, and on this it has built DEEP Agent, its document understanding AI platform.
This Series A is also the first outside assessment of how it has grown. That Korea Development Bank, a policy finance institution, increased its investment beyond its original plan shows that business results confirmed by revenue and customers formed the basis for the investment decision.
About DEEP Agent
Korea Deep Learning makes AI that does not stop at reading documents. DEEP Agent is a document-based AI agent that interprets a document's visual structure and linguistic meaning together to extract the key information, then carries out the whole of the document work, from classification and verification to the hand-off into business systems. Upload a contract and it summarizes the key clauses, flags the risky ones and goes on to register it in the ERP, and it reaches the field in two weeks on average with no separate training. It holds GS Certification Grade 1, Korea's national software quality certification, and is designated for priority purchase by public institutions. More information is available on the DEEP Agent 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.







