The Electronics and Telecommunications Research Institute has released TANGO, an open-source no-code platform designed to automate the path from data preparation and artificial intelligence model training to optimization and deployment. ETRI says the web-based tool is intended for practitioners who may not work with complex programming languages but need to adjust and deploy models within their own operational environments.
TANGO addresses a workflow that previously separated data collection, model development and software integration among different specialists. Its stated function is to support additional training of general-purpose foundation models, including DeepSeek and ChatGPT, with industry-specific data before reducing and deploying them to a target device or system. The platform does not remove the need for suitable data or technical oversight, but it brings the main stages into one development environment.
The project has expanded from TANGO1, which focused on computer vision, to TANGO2, which adds generative artificial intelligence and large language model operations. The core solutions are available through GitHub and can be installed on internal corporate servers, cloud infrastructure or on-device environments.
ETRI is testing the platform in several fields. Abeonotics is applying combined sensor data to an autonomous navigation solution for shipbuilding and marine operations, while SureDataLab is using agricultural experts' knowledge in a generative AI service for farms. Seoul National University Hospital has used 173,000 chest CT images to develop automated report generation and is preparing a cardiovascular disease prediction demonstration with Bundang Seoul National University Hospital. Commercial technology transfers are also being pursued, while international standardization work for no-code LLM operations is under way at ITU-T, with final approval targeted for May 2027.
