Rules Engine

Deterministic if-then logic that flags cases matching a pattern someone has already described.

No AIestablishedrules

A rules engine encodes what an organisation already knows about a pattern of concern.

It is the right starting point whenever the underlying logic can be stated plainly and the cost of a missed edge case is tolerable.

Reach for it before reaching for a model, and keep it in place even after a model is added, since it gives a cheap, explainable floor that the model then builds on.

It stops being enough once the patterns worth catching are too numerous or too subtle to write down one at a time.

General strengths

• Every decision is explainable in terms a non-technical reviewer accepts.
• Behaviour changes immediately when a rule is edited.
• Requires no training data, so it works on day one.

General limitations

• Only catches patterns somebody already thought to write down.
• Rule sets accumulate over time and start to contradict each other.
• Fixed thresholds are easy to learn and route around once known.

Applied to business problems

Each of these is written for its own problem. The technique is shared, the solution is not.

• Threshold and typology rulesDeterministic rules that flag wire, ACH, and cash activity matching known laundering typologies such as structuring and rapid layering.Transaction Monitoring
• Eligibility, authorisation, and claim edit checks before submissionDeterministic checks at registration, scheduling, and claim release that confirm active coverage, match each service to the prior authorisation its insurer requires, and stop claims with missing or invalid data before they reach the insurer.Claims Denied by Insurers
• Denial work queues routed by root causeRules that route each denied claim to the staff who know that insurer and that kind of denial, record the step that caused it, and send recurring causes back to the patient access, coding, or documentation team that owns them.Claims Denied by Insurers
• Degree map, registration, and balance alerts routed to a named adviserFixed alerts that fire when a student registers off their degree map, earns a low grade in a course that predicts success in their major, has not registered for next term, or is about to be dropped over a small balance, each sent to the student's own adviser to act on within a set time.Students Leaving Before They Finish Their Degree
• Intake red-flag rulesDeterministic checks applied at first notification of loss that flag a claim for review before payment.Claims Fraud Detection
• Condition threshold alarmsFixed alarms on individual sensor readings, such as vibration RMS or bearing temperature, that flag an asset once it crosses a set point.Unplanned Equipment Downtime
• Automated overpayment and fraud decisionsRaising a debt, suspending a payment, or recording a fraud finding automatically from a data match or a risk score without a caseworker deciding, described here because it has repeatedly harmed claimants and is the approach this playbook recommends against.Benefit Overpayments From Fraud and Error
• Earnings and circumstance data matchingMatching each claim against authoritative records other bodies hold, such as employer-reported pay, tax records, and death and prison records, using same-period figures to adjust awards with a right to dispute, and sending every retrospective or uncertain mismatch to a caseworker who asks the claimant first.Benefit Overpayments From Fraud and Error
• Reorder points and service level safety stockPer product and store reorder points and order-up-to levels that turn a demand forecast, the delivery lead time, and a planner's service level target into an order quantity.Empty Shelves and Excess Stock
• Ticket routing, macros, and a maintained knowledge baseStructured ticket forms, routing rules by product area, plan, and urgency, saved replies for known issues, and help centre articles that agents capture and update as part of solving tickets, so each ticket reaches the right person with a current answer.Support Queues Growing Faster Than the Team
• Feeder and transformer energy balanceComparing the energy measured entering each feeder or distribution transformer with the sum of the customer meters it supplies, so revenue protection teams work in the network areas where unbilled energy is concentrated.Electricity Theft and Non-Technical Losses
• Tamper-resistant metering for the largest customersFitting the small group of industrial and commercial customers that produce most of a utility's revenue with remotely read, tamper-resistant meters and checking their readings and events every day, before spending effort on the mass of small customers.Electricity Theft and Non-Technical Losses