South KoreaRetail and e-commerce2026 – ongoing

Cafe24

AI agents for merchant operations and promotions on Cafe24 PRO

Cafe24 cut a marketplace operations task from over 3 hours to under 5 minutes with AI agents on Databricks

WorkingVendor case studyVendor ↗

Problem

Cafe24 runs an online shopping mall platform with more than 2 million brands. It manages more than 300GB of log data a day across over 700 tables. Merchants needed faster analysis and less repetitive store work.

Approach

Cafe24 used Databricks and its Agent Bricks multi-agent framework, which splits complex analysis across specialized AI agents. The agents support its Cafe24 PRO operations service. Promotions were targeted by a customer's conversion stage, with cart reminders, revisit triggers and benefit timing. The company also planned automatic model drift detection and retraining.

Results

64%increase in average order value for Cafe24 PRO users supported by DatabricksVendor case study[1]

24%increase in total revenue over six monthsVendor case study[1]

under 5 minutestime for a marketplace operations task that took over 3 hours beforeVendor case study[1]

up to 100repetitive daily tasks per single-store operator now handled automaticallyVendor case study[1]

Takeaway

Most of the reported gains came from routine merchant work that agents took over and from better-timed promotions.

Sources

Driving growth for modern e-commerce brands
Databricks6 Mar 2026Vendor publication

Source summaries

[1]Databricks6 Mar 2026

Databricks describes how Cafe24 used Agent Bricks and targeted promotions for its Cafe24 PRO service. It reports a 64% rise in average order value, 36% more new customers and 24% more revenue over six months. A marketplace task fell from over 3 hours to under 5 minutes. The company did not say how these figures were measured.

Full details are in the original sources.

Added 4 Nov 2026 · Updated 4 Nov 2026