The Electronics and Telecommunications Research Institute (ETRI) has unveiled an integrated development tool that enables people who are not AI experts to develop AI models needed at industrial sites.

ETRI said on Sunday it released the no-code AI integrated development tool, Tango (TANGO), as open source.

Tango automates the entire development process, from training and optimisation to deployment, for AI models. Users can select and run needed functions in a web browser, which ETRI said can reduce the time required for AI development and ease the burden of securing specialist personnel.

ETRI has developed the Tango technology for five years since 2021. After Tango1, which supports AI development in computer vision fields such as video recognition, it expanded functions to Tango2, which supports generative AI and large language models (LLMs).

Tango2 supports LLM training, management and service application using industry-specific data. It can optimise and deploy developed AI models for different environments, including cloud, companies' internal servers and individual devices. The research team released LLMOps, a generative AI development and operations solution that is a core technology of Tango2, on GitHub.

ETRI is also pushing field applications. In shipbuilding and marine industries, AVENOTICS is demonstrating an autonomous navigation solution that supports safe ship operations by analysing sensor information with AI. In agriculture, Sure Data Lab built a generative AI service that learned agricultural experts' experience and knowledge for farms. In healthcare, Seoul National University Hospital developed a technology that automatically generates medical imaging reports using chest computed tomography (CT) images and report data.

ETRI is also proceeding with technology transfers to companies. Acryl linked the Tango LLMOps development tool to the development of its commercial product, Jonathan. Persona AI received a transfer of Tango technology in April and is strengthening LLMOps capabilities for its AI agent platform, Gen Station.

ETRI is also pursuing international standardisation. It is proceeding with international standardisation of no-code LLMOps technology at the International Telecommunication Union Telecommunication Standardization Sector (ITU-T).

ETRI will hold the fifth Tango open seminar on Nov. 11 at the Korea Science and Technology Center in Seoul's Gangnam district to share Tango2's key functions, development results and field application cases.

Cho Chang-sik (조창식), a principal researcher at ETRI, said Tango is a technology that supports people who are not AI experts in directly developing AI services needed at industrial sites. He said ETRI will continue to disclose development results and field application experience and provide an AI development environment that companies can use in practice.