AI full-stack company Elice Group, which provides AI infrastructure, cloud and industry-specific solutions, said on Wednesday it unveiled Helpy Document Vision, an AI document analysis solution that analyses complex-form documents on its own and converts them into structured data.

The company said Helpy Document Vision can automatically analyse not only paragraphs in documents but also various visual elements such as tables, charts, formulas and images. Helpy Table Vision is strong in analysing complex table data used in industrial settings such as large manufacturing companies. Helpy Document Vision can be quickly optimised for specialised documents by domain, including finance, healthcare and law, by organically combining in-house data and training infrastructure.

Elice Group plans to develop the vision language model (VLM) technology used this time beyond document recognition into a vision-language-action (VLA) model that can also make situational judgments and carry out actions.

VLA is a technology that supports AI in precisely controlling robots or machines in real industrial settings by looking at complex manuals or drawings. Soo-in Kim (김수인), CRO at Elice Group, said, "Through Elice's AI document analysis solution, we will help companies experience tangible innovations in work automation by converting complex documents previously handled manually into high-quality data." Kim added, "We will lead the era of 'physical AI' by advancing beyond VLM technology that simply reads documents to VLA technology that leads to real actions, directly solving problems at industrial sites."