Digital Presso, a company specializing in AI-based field data automation solutions, has established “AI Presso Lab,” an AI research platform that utilizes various conversational data generated in industrial settings, and is now working to advance domain-specific AI models.

AI Presso Lab is a platform for researching and validating the latest AI technologies—such as Vision AI, LLM, and STT—based on unstructured data (including audio, video, and images) generated in industrial settings. In particular, it focuses on experimenting with technologies using real-world field data and developing models tailored to industrial environments.

Digital Presso is securing call (STT) data from various industries through Proof of Concept (PoC) projects and collaborations with telecommunications and construction companies. Based on this data, the company is generating domain-specific synthetic data to enhance not only model performance but also practical applicability in real-world settings.

The AI models developed through this research aim to go beyond simple data analysis to automate a series of processes—ranging from the automatic generation of checklists based solely on call records to image classification and verification, report generation, and payment integration.

A Digital Presso representative stated, “AI Presso Lab aims to go beyond mere technical research to create AI models that can be practically applied in industrial settings,” adding, “We plan to further advance AI technology applicable to various industrial sectors, such as construction and manufacturing, through continuous learning and validation based on field data.”

Currently, AI Presso Lab is in the internal research and validation phase, with plans to expand its scope of application to various industrial sites in the future.