Introducing Fortify360 AI 2.0: Real-Time Risk Scoring for High-Velocity Compliance

For the past 2.5 years, Raj and I have been developing a new kind of risk model at Keyrails, specifically designed for the complex, high-stakes realities of onboarding medium- and high-risk clients across fragmented markets. We call it Fortify360.

At its core, Fortify360 is a proprietary quantitative risk model built on the Analytical Hierarchy Process (AHP). It dynamically calculates both inherent and residual risk, factoring in not only traditional AMLand risk pillars but also the operational and transactional nuances of our cross-border business. It gives us a real-time, model-driven view of how any client — whether a logistics firm in Nigeria using traditional banking rails or a UK-based PSP handling third-party stablecoin payouts — impacts our overall risk exposure.

The problem was scale. The model worked, but updating it meant hours of internal computation at a time when we were growing rapidly.

‍That’s now changed.

Meet Fortify360 AI 2.0

‍Together with our friends at AiPrise, we’ve been working on baking the Fortify360 risk assessment model into a custom model that has been built by AiPrise for quite some time which Keyrails is now expanding upon the model internally.

We can now assess the risk profile of any client during onboarding in near real-time.

From document tampering detection and metadata validation to source-of-funds analysis, sanctions screening, and false positive reduction, Fortify360 AI 2.0 evaluates four times more parameters than its predecessor. It delivers consistent, explainable risk scores in minutes rather than days, enabling our compliance team to make fast and informed decisions.

‍Why this matters

‍Our core market is international trade, particularly transactions involving at least one emerging market. That puts us face-to-face with Trade-Based Money Laundering , a form of laundering closely tied to illicit arms trafficking, drug trafficking, and sanctions evasion. Managing this risk demands more than checkbox compliance.

‍Fortify360 AI 2.0 provides us with the precision, speed, and auditability necessary to operate effectively in these environments.

What’s next

‍We’re integrating Fortify360 AI with our internal operations AI, which has been trained over the past six months on our workflow data. This will create a unified layer between PayOps and Compliance. Every transaction can be automatically cross-checked against the client’s risk profile. Smart red flags are raised without slowing down operations.

This is what AI-native compliance looks like — a fully connected system from onboarding through transaction monitoring, built for speed, scale, and explainability.

‍We’re not waiting for the future of compliance. We’re building it.

‍Stay tuned.

Berhan Kongel