Evaluator ATM Skimmer Detection

Evaluator turns short clips from the cameras you already have into an early warning against card skimming at ATMs, point-of-sale terminals, and fuel pumps.

WHAT IT DOES

Evaluator's AI analyzes short event clips from your ATM, point-of-sale, and fuel-pump cameras and flags card-reader tampering: someone fitting a skimmer, overlay, insert, or other foreign device to the machine, or repeatedly manipulating the card reader. Each result arrives within seconds of the clip reaching the running engine, with the model's reasoning and a link to the clip, so your team can check it and respond before cards are compromised.

WHO IT'S FOR

Built for retail and commercial banks, independent ATM deployers (IADs), ATM hardware operators, and retail, forecourt and self-service operators running point-of-sale and unattended payment terminals. It scales from a pilot on a few machines to large fleets and fits the workflows of fraud prevention, physical security, and operations teams.

THE PROBLEM IT SOLVES

Skimming remains a persistent and costly source of fraud and reputational risk. Traditional monitoring depends on manual review or after-the-fact reporting, leaving a window in which compromised machines keep operating. Adding sensors across a large fleet is expensive and slow. Evaluator closes that gap using the camera infrastructure already in place, with no sensors to fit on the machines.

HOW IT DEPLOYS

Evaluator is delivered as an Azure Managed Application that deploys into your own Azure subscription. Clips and detection results stay there, supporting strict data residency, privacy, and regulatory requirements. Every installation runs its AI engine on an Azure confidential VM with an NVIDIA H100 GPU in confidential mode: the model and the video frames it analyzes are processed in encrypted memory. You start and stop the GPU from the dashboard, and there is no GPU charge while it is stopped. Alerts go to your existing tools (webhook, CEF syslog, Microsoft Sentinel or Splunk) rather than replacing them.

KEY BENEFITS

- Flags card-reader tampering at ATMs, point-of-sale terminals and fuel pumps within seconds of each clip

- Works with clips from the cameras you already have; no sensors on the machines

- Runs in your Azure subscription on confidential computing; clips and results stay in your environment

- Alerts into your existing SIEM and monitoring tools

- Microsoft Entra sign-in with five roles, and an audit log of every clip viewed

- Supports bank security, privacy and model-risk reviews with published accuracy figures and a security disclosure

- Scales from a pilot to enterprise fleets