Glemad EditorialStories2 min read
Nigerian Bank Moves ATM Operations to Predictive Failure Intelligence With Ollandi
A Nigerian commercial bank uses Ollandi for ATM to open cases before service thresholds are breached and to separate terminal faults from switch, power and carrier failures.
Summary
A Nigerian commercial bank deployed Ollandi for ATM to move its ATM operations from reacting to customer complaints to acting on forecast risk. Developing faults are now opened as cases before they affect customers, and field work is directed only to terminals that need it.
Before Ollandi
- Most failures were reported by customers, not by monitoring.
- Terminals stopped dispensing while their monitoring status remained green.
- Switch, power and carrier problems generated floods of individual alerts.
- Engineers were sent to terminals that had no fault.
- Silent data feeds were counted as healthy terminals.
- Failed anti-skimming modules and cameras received low priority.
After Ollandi
- Developing faults are opened as forecast-risk cases before service thresholds are breached.
- Availability, dispensing, communications, cash, hardware, software and security are tracked as separate health dimensions.
- Terminals affected by a switch link are grouped into one case explained by the upstream cause.
- Generator and UPS operation is recorded as normal; cases open only when backup capacity is at risk.
- Terminals behind a silent feed are shown as stale or unknown, never healthy.
- Offline security controls are routed to the security team as exposure, with full evidence.
Deployment
- Product: Ollandi for ATM.
- Data sources: ATM monitoring, switch state, device events, site power and carrier status, security signals. All access is read-only.
- Operating control: Observe & Notify. Email escalation was enabled after the bank validated the workflow and two administrators approved it.
- Ollandi does not reset terminals or dispatch engineers. Repairs remain with the accountable team or vendor.
By the numbers
Measured across Glemad's production deployments and published in the the earlier model research research paper:
- Threat coverage: 40 to 60% more multi-step attack chains identified than siloed detection tools.
- False positive rate: 1.2%, against an industry average of 23.5%.
These figures come from environments with mature telemetry and defined policy baselines. Results depend on the quality of the telemetry available.
