> ## Content Index
> Fetch the complete content index at: https://www.ai-360.online/llms.txt
> Use this file to discover other available public pages before exploring further.

# LG Uplus uses Arize AX to evaluate its contact-centre AI agents
- URL: https://www.ai-360.online/lg-uplus-builds-better-ai-customer-service-agents-with-arize-ax-evaluation/
- Published: 2026-09-01T12:00:00.000Z
- Updated: 2026-10-31T22:24:10.000Z
- Description: LG Uplus builds evaluation datasets first and feeds production traces into Arize AX to improve contact-centre AI agents serving more than 30 million users.
- Author: AI-360
- Tags: 1. Enterprise AI, News, #company-arize-ai, Enterprise AI, Customer

[LG Uplus](https://arize.com/?ref=ai-360.online) has adopted evaluation-driven development with [Arize](https://www.ai-360.online/arize-ai/) AX to improve the AI agents in its contact centre, which serves more than 30 million subscribers, [according to an Arize customer case study published on 17 August 2026](https://arize.com/customers/how-lg-uplus-is-building-better-agents-with-evaluation-driven-development-and-arize-ax/?ref=ai-360.online).

The South Korean operator's AI Contact Center team builds systems that route and process customer calls, assist service representatives during conversations and automate post-call work. The team now builds evaluation datasets first and refines them as the agent improves, feeding production traces, user feedback and expert review into a closed loop.

“The development process itself changed,” said Minkyu Ha, a senior software engineering manager at LG Uplus. “Once the traces are collected, the evaluation cycle becomes automated.”

Ha said assessing intermediate steps had been the hardest part. “The most important part of an agent flow is selecting tools and optimizing routing paths. Before, evaluating those intermediate processes was difficult. After introducing Arize, that improved significantly.”

---