Launching the Coris Intelligence Score
Scoring Every Merchant, Every Day
At Coris, we’re building infrastructure to help risk teams scale smarter—without getting buried in rules and false positives.
Today, we’re introducing “Coris Intelligence”, an AI-powered scoring system that continuously analyzes each merchant in your portfolio and assigns a dynamic risk score. It’s already helping teams cut down false positives by up to 60%, and we're currently rolling it out in beta to existing Coris customers.
The model is trained on your platform’s data, the output is calibrated to your portfolio—not a generic risk profile.
Rules Alone Don’t Scale
One of the most common questions we hear from risk teams is: “Which merchants should we be looking at today?”
For most platforms, the answer is: write more rules. Flag chargeback spikes. Track refund volume. Look for unusual payment activity.
The challenge? These rules tend to be overly broad and rigid. They catch the obvious cases—but they also pull in a long list of good merchants that don’t require review.
In fact, across most teams we work with, 70–80% of alerts end up being false positives. That means lost time, overloaded queues, and attention pulled away from actual threats.
Enter Coris Intelligence Model
To solve this, we created the Coris Intelligence Score—an AI model that evaluates risk at the merchant level, continuously.
On a continuous basis, the model:
- Processes recent merchant behavior
- Compares it against historical trends
- Benchmarks against the broader portfolio
- Assigns a Coris Intelligence Score from 0 to 100
A higher score means higher relative risk—and every score comes with a breakdown of the top reasons why the merchant was scored that way.
This means teams don’t just get alerts—they get context.
What goes into the model
The model looks at multiple dimensions of activity on a continuous basis, including:
- Disputes, refunds, and chargebacks
- Shifts in transaction patterns
- Declines and payer-level anomalies
- Merchant metadata (age, volume, behavior)
- Portfolio-wide baselining and outlier detection
What Makes It Unique
- Built per platform
Each model is customer-specific, ensuring alignment with your merchant segments, thresholds, and tolerance for risk. - Explainable scores
The Coris Intelligence Score isn’t a black box. Every alert includes the key reasons and relative weights behind the score. - Smarter prioritization
By applying deeper merchant-level signal analysis, teams are seeing 50–60% fewer false positives—and spending more time on the alerts that actually matter.
How Customers Are Using It
The Coris Intelligence Model is currently available in beta for Coris customers. It can be used to:
- Power merchant-level risk alerts
- Replace or enhance rule-based triggers
- Triage investigation queues
- Automate reviews for low-risk merchants
If you're using the Coris Risk Platform, please reach out to see if we can have the model enabled—then you can begin leveraging the Coris Intelligence Score right away.
What’s Coming Next
This is the first step in a broader roadmap.
We’re actively working on:
- Analyst feedback loops (so risk teams can provide direct feedback in natural language)
- Continuous model tuning based on outcomes
- Advanced segmentation across merchant cohorts, types and geographies
As with all Coris models, these updates will happen behind the scenes—without requiring manual tuning on your end.
Ready to scale merchant monitoring - without scaling the noise?
The Merchant Risk Model helps teams spend less time writing rules and more time making decisions. It’s one more step toward the modern risk infrastructure we’re building at Coris.
.svg)
Related Resources
.png)
.png)
.png)
.png)
.png)
.png)
.svg)
