South KoreaRetail and e-commerce2026 – ongoingGolfzon Commerce
Multi-agent AI shopping agent that recommends golf clubs and balls from swing data
Golfzon Commerce put a multi-agent AI fitting service on its Golfping online mall in May 2026
Problem
Golfzon Commerce sells golf equipment through its online mall Golfping and more than 110 Golfzon Market stores. Customers choosing clubs and balls online had no personal fitting to guide them.
Approach
Megazone Cloud built the system on Amazon Bedrock AgentCore with Claude Sonnet 4.5. It has more than 10 agents in two groups. One group profiles the golfer from swing data. The other infers suitable equipment from that profile. The models draw on more than 500,000 fitting records from Golfzon screen golf venues. The service went live on 7 May 2026 and suggests three clubs per customer with reasons and expected effects, plus matching golf balls.
Results
Takeaway
The retailer used fitting data it already held from its venues as the base for online product advice. No sales or conversion results had been published.
Categories
- Industry
- Retail and e-commerce
- Business function
- Marketing and sales
- Innovation type
- Product and service innovation
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
Reports Megazone Cloud's announcement of an AI shopping agent for Golfzon Commerce. It describes the two agent groups, the Bedrock AgentCore setup, the 500,000+ fitting records and the 7 May 2026 launch. It gives no sales or usage results.
Full details are in the original sources.