Musinsa MATCH Console & MAI: Connecting Data Exploration to Execution

Musinsa's Core-E Personalization team unveiled the operations console MATCH Console and the AI assistant MAI for their personalization system MATCH. The core goal is to reduce inefficiencies in data verification and campaign setup that previously occurred between marketers, analysts, and developers, allowing operators to focus on their primary task: 'decision-making'.

MATCH Console connects three stages—data exploration, audience definition, and execution channel integration—into a single flow.

• Data Explorer: Converts technical database column names into business language and verifies the current data collection status (Active/Inactive) in real time.
• Audience: Defines target audiences by combining conditions like natural language sentences without SQL syntax, and allows pre-verification of the actual customer profiles included.
• Activation: Passes defined audiences to CRM campaigns or the Audience API, supporting flexible usage methods such as MATCH handling final targeting or providing information only to the service side.

MAI is an AI assistant based on a ReAct (Reasoning + Acting) agent architecture. When operators ask questions in natural language, MAI explores relevant tables and schemas, generates and executes SQL queries, and then summarizes the results. This reduces the data verification process, which previously relied on analysts, to under one minute within the console, and reduces users' cognitive load through recommended questions.