Korea’s Upstage Lands $45M from KDB, Amazon, AMD to Scale Enterprise-Focused AI

Upstage, a South Korea-founded AI company with operations in San Jose, has raised $45 million in a Series B bridge round to strengthen its enterprise-focused generative AI offerings. The funding was led by Korea Development Bank (KDB) with participation from Amazon and Advanced Micro Devices (AMD), bringing the company’s total capital raised to $157 million.

The fresh investment will support the development of Upstage’s next-generation Solar large language model, expansion of its document AI suite, and accelerate go-to-market efforts in the U.S. and Asia-Pacific markets. Upstage’s document intelligence technology converts unstructured files, including scanned paper documents, into structured formats with optical character recognition (OCR) accuracy of up to 99%, enabling enterprises to extract and apply critical data more efficiently. 

Upstage plans to use the fresh funding to advance its Solar language model line, strengthen its document AI offerings, and push further into the U.S. and Asia-Pacific markets. As part of this effort, the company has expanded its partnership with Amazon Web Services (AWS), designating it as its primary cloud provider. Through AWS, Upstage is training and deploying its foundation models using Inferentia, Trainium, and SageMaker.

The funding follows a $72 million Series B round in 2024, led by Korean investors such as SK Networks and Shinhan Venture Investment, bringing Upstage’s total raised to $157 million. The company said the additional resources will support R&D to advance Solar’s enterprise use cases in areas like document parsing, task automation, and workflow optimization.

Upstage’s strategy is the Document Intelligence suite, which includes tools for converting unstructured files into structured, LLM-ready formats and extracting key data points with accuracy rates above 95%. Unlike traditional OCR systems that capture static fields, Upstage’s platform interprets documents holistically — accounting for layout, meaning, and intent — to generate structured outputs suited for real-world business processes.

The technology is being applied to industries burdened by paper-intensive processes, especially insurance, where tasks like manual claims handling and broker submissions slow down operations. In the U.S. alone, claims adjudication costs the insurance sector an estimated $25.7 billion each year, with $18 billion linked to overturned denials and avoidable reviews. Upstage aims to tackle these inefficiencies head-on.

“Our collaboration with AWS brings scale, credibility, and deep technical alignment at a time when enterprises need production-ready AI,” said Sung Kim, co-founder and CEO of Upstage. “The next phase of AI is about precision and real-world complexity — and that’s exactly what we’ve built into our platform.”

Upstage’s tools are already in production with major clients, including Samsung and leading Korean insurance providers, while adoption is expanding across the U.S. insurance market and into sectors such as healthcare, manufacturing, and government services. The company also offers its models on AWS marketplaces, making them more accessible to enterprises seeking ready-to-use AI agents for document-heavy workflows.

The Series B bridge will also fund sales and marketing hires to support international growth. Recently named to the CB Insights AI 100 list, Upstage said the recognition reinforces its momentum in global markets and its ambition to become a leading player in enterprise generative AI.