First Community Bank and Trust moved from threshold-based sampling to consistent review of every check. Socratix now evaluates each item, auto-clears 97.5%, and gives analysts clear evidence for the checks that need human judgment.
We went from reviewing a fraction of our checks to reviewing all of them, while cutting our review time by more than half in the process.
A century of community banking
First Community Bank and Trust has served the Beecher and Peotone communities in Illinois for more than 100 years. Today, with $200 million in assets, the bank combines personalized community banking with a willingness to embrace new technology.
That mindset is reflected in one of its core values, “Change and Innovation,” and in its decision to bring AI into one of the bank's most critical operational workflows: check-fraud review.
The limits of threshold-based review
Check washing, signature mismatches, and MICR amount discrepancies can be difficult to identify through manual inspection alone—especially when reviewers are responsible for thousands of checks each month.
Like many community banks, First Community relied on a flat dollar-amount threshold to decide which checks warranted a closer look. The team reviewed roughly 18% of business checks and just 3% of consumer checks, leaving the remainder to clear without a second review.
The bank recognized that protecting customers while maintaining operational efficiency required a smarter and more consistent approach.
A partner who understood community banking
First Community was not simply evaluating another fraud-detection tool. The bank wanted a partner that understood how community banks operate and would collaborate closely on an effective solution.
The pilot quickly demonstrated value beyond the bank's initial expectations. It identified potential fraud while surfacing other issues that had previously gone undetected, including missing signatures, mismatched written and numeric amounts, inconsistent handwriting, and other check irregularities.
You caught so many things that are wrong with the checks that are being cleared today. They might not be fraud, but they're still wrong. Customers expect that we're paying attention to the checks they're writing. If it's wrong, and it's a material dollar amount, we can get in front of it and engage the customer.
Results with Socratix
First Community dramatically reduced the manual effort required to review checks while improving consistency across the entire workflow.
of checks evaluated
of checks auto-cleared
less manual review time
- Reduced the time spent reviewing checks manually every day by 65%.
- Coverage expanded from 18% of business checks and 3% of consumer checks to every check
- Every item is evaluated against historical customer behavior and account activity
- Potential signature discrepancies, MICR mismatches, and signs of check washing are identified automatically
- Flagged checks arrive with supporting evidence and recommended next steps
I would say the time savings is a minimum of 65%, and more importantly, the time our team does spend is now focused on items that have a specific reason for being looked at.
Helping analysts focus where it matters most
Before Socratix, employees pulled check images from cold storage, selected comparison ranges, reviewed signatures and alterations, and rebuilt their baseline for each investigation. Decisions could vary with each reviewer's individual judgment.
Now, the check image, historical comparison, and narrative explanation arrive together in one view. Every flagged item has a consistent, definitive reason, while analysts retain the choice to follow or override the recommendation.
It's really surprising how well it reads the imagery of the entire check. The handwriting comparisons by themselves are magical, and it gives you a narrative. I'm struck by descriptions like “hard pen stroke looping to the right,” versus just saying “it's different.” The system gives you a definitive reason as to why the item was flagged for additional review.
By handling the initial investigation and auto-clearing low-risk checks, Socratix gives the team more time for the smaller number of high-risk alerts that require human judgment and expertise.
The impact extends to customer service. In one case, a check written for $1,100 had cleared for $2,200. The analyst contacted the customer, who was able to address the discrepancy immediately with the vendor. First Community now reaches out whenever a flagged check appears suspect, even when the issue is not fraud.
Looking ahead
First Community continues to grow its partnership with Socratix and identify other areas where AI can support the bank's processes. As fraud techniques evolve, the team now has a scalable workflow that reviews every check consistently while keeping people focused on the cases that require their judgment.
In golf, we call it ham and egging it. With this solution, the bank does what it's good at and Socratix does what it's good at. That's what a real partnership should be. We had ideas, Socratix had ideas, and together we're building something that really works! I'm excited to see what we can do next.