Client story · Retail & procurement

AI purchase negotiation trained on tens of thousands of real deals

Working with KT Commerce, AIO2O trained a negotiation model on real transaction data so that an AI agent can take over the price negotiations that repeat across business-to-business deals.

Client: KT Commerce
Industry: Retail & procurement
Solution: NEGOTIUM
Period: 2025

Challenge

Smaller deals went through without negotiation

Price negotiations in business-to-business deals become repetitive and inefficient as the number of items and suppliers grows. Buyers cannot negotiate every deal themselves, so smaller deals often go through without any negotiation.

Approach

A negotiation agent trained on real transaction data

Designed to automate and optimize the repetitive, inefficient price negotiations in business-to-business deals with AI.
In June 2025, AIO2O signed an agreement with KT Commerce and the Korea University HI-AI Institute to develop an AI negotiation agent.
The agent analyzes past deals and market prices to set a target price, then negotiates with suppliers in conversation.

Results

What changed

The negotiation model was trained on tens of thousands of real KT Commerce transactions.
This negotiation solution won the ITU AI for Good Innovate for Impact Awards 2025, which drew entries from 320 teams worldwide.

Video