Building an AI Agent to detect risky merchants at scale

The client

A global payment processor helping financial institutions launch and scale payment solutions efficiently was facing rising fraud threats. To stay ahead, the client recognised the need to strengthen their fraud prevention capabilities-particularly in how they detected and mitigated risky merchants.

‍

The challenge

The client wanted to automate as many manual fraud processes as possible. One major challenge was identifying and blocking high-risk merchants. Their existing approach was time-consuming, inconsistent, and costly-often allowing fraud to slip through the cracks.

‍

Key issues:

Heavy manual workload: Fraud analysts had to manually analyse transaction data, increasing the risk of missed fraud.

Poor scalability: As business and transaction volumes grew, so did the demand for more analysts.

Slow detection: Merchant risk assessments took weeks, giving fraudulent actors more time to operate undetected.

‍

‍

The solution: Risky Merchant Agent

Fortify partnered with the client to develop an AI agent that automates the identification of high-risk merchants and blocks suspicious transactions-using the client's existing tech stack and data sources.

‍

Key features:

Automated risk assessment: The agent flagged merchants based on indicators such as invalid cards, expired cards, and chargebacks.

Historical behaviour analysis: The model looked at merchant activity over three to six months to ensure accurate trend detection.

Impact-aware mitigation: The agent assessed the risk of customer disruption before blocking, protecting legitimate merchants from unnecessary penalties.

‍

Key takeaways

Automated efficiency: Risk detection automation boosted both fraud prevention outcomes and operational productivity.

Faster impact: Real-time merchant analysis reduced false positives and sped up fraud response.

Plug-and-play integration: The agent seamlessly integrated into the client's existing tools-no new infrastructure required.

‍