SK Broadband has built an environment where any employee can directly develop AI agents and apply them to work, and has begun upgrading network quality management.
SK Broadband said on the 2nd that its in-house network organization and the AT·DT Center jointly built the Playground platform in Feb., which comes with network data analysis and coding support functions. Previously, complex procedures such as applying for server access rights, building a development environment, and installing libraries were required for data analysis or automation, but the platform cut the development preparation period—which used to take more than two months—to about 5 minutes.
Playground is linked to the location-based in-house data analysis system LDAS. Employees can create AI apps using network equipment, quality and traffic data, and customer experience indicators. About 600 AI apps are currently under development and operation, and among them, a little over 30 have been applied in the field as AI agents that assess situations on their own and carry out actions.
A representative example is the AI control and diagnostics agent C-One. Based on customer experience indicators, C-One automatically detects signs of anomalies on the fixed-line network and suggests causes and inspection priorities. It allows users to check areas or buildings with suspected quality degradation on a map, and when AI analysis is run, it aggregates changes in quality scores, outage reports, and optical signal strength to propose the problem section and a mitigation plan.
SK Broadband plans to upgrade C-One into an autonomous recovery agent that independently performs everything from fault detection to handling and restoration.