Technology & Research
Support that's easier to trust.
Fraud, waste and abuse in publicly funded services (Medicaid home- and community-based services) hurt the people who rely on them most. We're researching better ways to verify that support really happened — without burdening honest workers or tracking people more than needed.
Why it matters
In FY 2025, the U.S. Department of Health and Human Services Office of Inspector General reported 326 fraud convictions involving personal care services attendants — more than any other individually identified provider type. A 2026 GAO review also found fraud-risk assessment gaps across major state-administered programs.
The problem with one-size-fits-all checks
Today's visit-verification tools check GPS, time entries and authorizations. But a changed destination, a missing GPS signal or no client confirmation is often perfectly normal for a community outing, help at home or respite. Rigid rules flag honest work and can still miss real problems.
Our research approach
Service-aware verification
Checks are based on what was actually authorized. A community outing, help at home and respite each look different — so they shouldn't be judged by the same rigid rules.
Honest about uncertainty
Records can be missing, conflicting or wrong for ordinary reasons. We weigh how reliable each piece of evidence is instead of treating gaps as proof of wrongdoing.
Privacy first
An extra check only happens when it would really help. No continuous location tracking — just the fewest, clearly disclosed checks needed.
How we'll test it
Using synthetic records that include legitimate schedule changes, GPS errors, lost connections, duplicate claims and manipulated records, we'll compare our approach against today's rule-based checks and standard anomaly detection. We'll measure:
- — How well real billing problems are caught
- — How often honest work is wrongly flagged
- — Whether the system's confidence matches reality
- — How little location data is collected per visit
Our principles
- Automated flags lead to a human review — never a finding of fraud.
- A client not confirming a visit is never treated as misconduct.
- Research uses synthetic (fictional) records — never real client information.
- Honest workers shouldn't carry the burden of catching dishonest ones.
Where we are
Solenz LLC, the company behind Everyday Independence, has submitted a Project Pitch to the National Science Foundation's SBIR/STTR program (Digital Health). This is early, ongoing research: the approach and its advantages remain to be demonstrated, and no funding has been awarded at this stage.