- Customer Stories
- Nevada SNAP
Nevada SNAP achieved this using • SAS® Payment Integrity for Food Assistance on SAS® Viya® deployed with SAS® Managed Cloud Services
For families relying on food assistance, an incorrect benefit amount can have consequences long after the payment arrives. Too little can mean choosing between groceries and another essential bill. Too much can mean a debt the family must later repay. Getting the amount right matters to the household budget.
Helping families receive accurate, timely benefits is central to the work of the Nevada Division of Social Services. The agency administers the Supplemental Nutrition Assistance Program, or SNAP, which helps eligible individuals and families buy food.
In Nevada, nearly 400,000 people receive SNAP assistance, which total about $71 million in benefits each month. The agency also handles roughly 26,000 new applications and 36,000 renewal applications monthly. At that scale, staff cannot manually review every case each month.
The challenge is knowing where to look first. Nevada uses SAS Payment Integrity for Food Assistance to analyze the full caseload with data and artificial intelligence, assign risk scores and guide staff to cases most likely to contain payment errors. That gives staff a starting point for review – not a substitute for their judgment.
Now we have a report that simply says, ‘Start here.’ We can begin with the highest-risk cases and work our way down.Kelly Cantrelle Nevada Division of Social Services
From a sample to the full caseload
“Sheer volume was the biggest barrier,” says Kelly Cantrelle of the Nevada Division of Social Services. Previously, the agency’s quality control team reviewed a random sample and used it to estimate the overall error rate. The sample helped measure overall payment accuracy, but it did not show staff which cases across the full caseload were most likely to contain errors.
The need to identify those cases has become more urgent. Cantrelle says Nevada has historically maintained a SNAP payment error rate between 6% and 7.99%. Under new federal requirements, remaining in that range could require the state to contribute close to $44 million a year toward benefit costs. That added expense could affect staffing, facilities and, ultimately, customer service.
For Nevada, reducing errors is not simply about avoiding a potential cost. It is about protecting the program’s ability to serve eligible families while balancing speed and accuracy.
“Nevada citizens who are eligible for SNAP are entitled to accurate benefits,” Cantrelle says. “Accurate, timely benefits are our number one goal.”
A practical starting point for every review
Nevada began taking action before the new requirements were finalized. The agency strengthened verification procedures, refreshed staff training and increased the frequency of customer interviews during the benefit renewal process. After receiving funding for technology initiatives, it added SAS to help extend those efforts across the entire SNAP caseload.
Each month, the SAS payment integrity solution analyzes case data and assigns risk scores so staff can prioritize cases most likely to contain errors. The analysis may flag a data-entry mistake, an unreported change or activity that warrants further investigation. Staff then review the findings to determine whether a correction or further action is needed.
“Before, we were searching, looking and hoping to find them,” Cantrelle says. “Now we have a report that simply says, ‘Start here.’ We can begin with the highest-risk cases and work our way down.”
The focused approach helps a small team review far more cases than would be possible through manual review alone. It also produced usable information quickly. Within five months of receiving funding, Nevada had its first SAS report and began putting the findings to work.
For Cantrelle, who has worked in the field for more than 26 years, that pace stood out. “I have never seen something move that quickly to the point where we had something in hand that we could use to help bring down our error rate,” she says.
Nevada SNAP – Facts & Figures
400,000
people receiving SNAP assistance
$71M
in SNAP benefits distributed monthly
$1.2M
in fiscal savings in the first six months
Turning findings into action
The reports gave Nevada concrete leads to investigate, from unusual shelter expenses to contact information shared across multiple cases.
“The SAS platform can look at the median shelter expense for a ZIP code area and easily identify a shelter expense that is much higher than the norm,” Cantrelle says. “If we see a case with $10,000 a month in rent instead of $1,000, we know that is probably a data-entry error.”
Other patterns may reveal connections across cases. “In one instance, 45 cases were using either the same phone number or the same email address,” Cantrelle says. Nevada’s team was able to refer the findings to its investigations and recovery specialists for further review.
Those examples helped change Cantrelle’s perspective on AI. Initially apprehensive about the technology, she came to see how it could analyze large volumes of data quickly while staff applied their program knowledge to verify the findings.
“We do not simply trust what is in front of us; we use it and verify it,” she says. “The technology and human expertise have to work hand in hand.” Caseworkers and program specialists remain responsible for reviewing the information, contacting customers when clarification is needed and making the final determination.
Early impact with more to come
Nevada is already seeing measurable results.
In the first six months of using SAS findings, Nevada reviewed 5,900 cases and estimates savings of more than $1.2 million. The agency closed 756 cases after reviewing the findings, following up with customers and evaluating information already on file. The work has also helped identify errors, misuse and suspected fraud.
The impact reaches beyond the agency’s balance sheet. Cantrelle notes that SNAP benefits flow through a broad food network that includes farmers, ranchers, transport providers, retailers and local communities. Improving payment accuracy helps direct public resources appropriately while supporting the people and economic activity connected to the program.
Over time, Nevada expects the monthly analysis to reveal recurring patterns that can inform staff training and process improvements. The agency can also request reports on emerging issues involving shelter costs, income or potential fraud.
Looking ahead, Nevada needs a solution that can keep pace with changing caseloads and federal requirements. Deployed with SAS Managed Cloud Services, the SAS solution can scale as those needs evolve – an important part of its long-term value.
“The SAS platform has the flexibility and adaptability to change with federal guidelines,” Cantrelle says. “We all know that the situation, rules and regulations we work with today will not necessarily be the same tomorrow.”
Keeping people at the center of AI
For other public agencies facing similar pressures, Cantrelle’s message is practical: Having the data is only the beginning. Nevada assigned a team to review its first report immediately and continues to use the findings to guide reviews. Staff verify the evidence and apply their program knowledge before making decisions.
“Do not be afraid of AI,” Cantrelle says. “AI is not replacing our staff, our expertise or our decision-making. We have the knowledge and historical background. AI gives us a platform that allows us to do far more than we could do manually with the people we have.”


