The Electronics and Telecommunications Research Institute (ETRI) contributes to enhancing the competitiveness of the domestic software industry by releasing no-code machine learning development tools.

ETRI announced that it will release the core technology of the no-code machine learning development tool (MLOps), developed with support from the Ministry of Science and ICT (hereinafter MSIT) and the Institute of Information and Communication Technology Planning and Evaluation (hereinafter IITP), as open source, and will hold a public seminar at the Science and Technology Hall on the 1st to expand the GitHub community.

Since 2021, ETRI researchers have been developing the TANGO framework, which enables users with limited expertise in artificial intelligence in industrial fields such as factories and healthcare to automatically generate neural networks based on no code and automate the deployment process, and have been releasing the core technology as open source since last year.

The Tango framework is a technology that automatically develops AI-powered application software and optimizes and deploys it for various device hardware environments, such as cloud, Kubernetes edge environments, and on-device.

The existing AI application software development method was structured so that domain experts handled data labeling, while software developers directly performed AI model development and training, as well as the installation and execution of the application software.

With the spread of artificial intelligence technology, the demand for software is rising across all industries, but AI and software specialists to meet this demand areIt is a sufficient situation.

Recently, in an effort to address these issues, research to automate the development and distribution of AI application software has begun, led by global companies such as Amazon, Google, and Microsoft; however, since they provide development environments tailored solely to their own service environments, there have been difficulties in supporting the diverse hardware of domestic industrial sites.

Reflecting such demands from domestic industrial sites, ETRI is developing neural network automation development algorithms optimized for object recognition.

In particular, it supports the optimization and automation of the entire process, including data labeling, AI model creation, AI training, and application software deployment, to enable practical application in industrial settings such as medical facilities and smart factories.

Cho Chang-sik, Head of the AI Computing System Software Research Lab at ETRI, said, “We plan to actively release the Tango framework and rapidly commercialize the technology through joint development in cooperation with industry, academia, and the community. Going forward, we will continue to release new versions of the source code on GitHub every six months and plan to hold a public seminar once a year in the second half to share the technology.”

Jang Moon-seok, SW PM at the Institute of Information & Communication Technology Planning & Evaluation (IITP), also stated, “Once the technology development of TANGO is completed, domestic cloud companies will be able to secure industrial competitiveness in the field of AI development tools, which is currently monopolized by foreign cloud services. ETRI’s knowledge and experience in neural network development will be of great help in enhancing the competitiveness of the domestic software industry.”