From Reactive to Proactive: Pharmacy Fraud Detection
Healthcare fraud drains billions of dollars from Medicare and Medicaid every year, and pharmacy and opioid-related fraud remains one of the hardest patterns to catch early. Congress funds prevention through the Healthcare Fraud and Abuse Control program, and the proposed Anti-Fraud Fund Act would add further funding through 2030. Even with more enforcement dollars, most fraud is still discovered only after the damage is already done, leaving state Medicaid programs reacting instead of preventing.
This showcase demonstrates a proactive, AI-powered fraud detection solution built for state pharmacy and opioid claims data. Instead of relying on manual reviews, the system continuously analyzes providers, claims, and risk signals across every detection period, scoring risk against 89 statistical fraud indicators. It explains why a provider was flagged by comparing them to county peers, their own prescribing history, and statistical outlier thresholds, then automatically assembles a complete case package with evidence ready for investigation.
The workflow is powered by Jarvis AI. Jarvis Chat gives investigation teams a natural language interface to ask questions, such as the overall risk picture for a county, and get evidence-backed answers pulled directly from live claims data. Jarvis Registry provides the governed layer behind the workflow, connecting agents, tools, and data sources with visibility, control, and traceability, turning complex fraud analytics into an actionable investigation workflow.
Medicare and Medicaid lose billions of dollars a year to fraud, waste, and abuse, including pharmacy and opioid-related fraud. Even with HCFAC funding and the proposed Anti-Fraud Fund Act, most fraud is still discovered only after the damage is already done.
The video introduces an AI-powered proactive fraud detection solution for state pharmacy and opioid claims data. It continuously analyzes providers, claims, and risk signals, scoring risk across 89 statistical fraud indicators instead of waiting for manual review.
A specific provider is marked critical after the system detects multiple risk signals at once, including an opioid prescribing rate 11.7x higher than the county average. The system also compares the provider against their own history, revealing a 19.7 percentage point increase over time.
Once a provider is flagged, the system automatically builds a complete case package containing risk factors, historical trends, statistical evidence, and peer deviation data, ready to support an investigation without manual assembly.
Investigation teams can ask plain-language questions, such as the overall risk picture for a county, instead of digging through spreadsheets. The system responds instantly with every flagged provider ranked by risk level and exposure from live claims data.
The workflow is powered by Jarvis AI, with Jarvis Chat providing the natural language interface and Jarvis Registry serving as the governed layer connecting agents, tools, and data sources with visibility, control, and traceability.
