Stop repeated notifications from creating duplicate work

A useful starting point

A workflow can receive the same event more than once. Record a stable event ID before creating an invoice, task or outbound message. A rule that only runs on approved items helps filter noise, but does not prove that each item is handled once.

A worked example

For a hypothetical order event, reserve order-123:confirmation before sending its confirmation. Store whether the send was accepted, rejected or uncertain. If a timeout follows transmission, stop for reconciliation instead of automatically sending another message.

Implementation and review

Ask your implementer how reservations survive restarts and concurrent runs. Test two simultaneous copies of an event, a crash after the external action and a later replay. Keep uncertain outcomes visible to an operator. Trigger conditions are a useful first filter, with durable duplicate controls added where side effects matter.

Trigger conditions can restrict when a flow starts; they do not by themselves guarantee duplicate prevention. Microsoft: trigger conditions

Measure the improvement

Track duplicate actions and reconciliation time during testing. Include the operator workload in assisted cost. This is reliability work, not a claim of perfect delivery. Compare implementation and storage costs with the real cost of duplicate actions in your process.

Discuss an implementation

This guide applies across Canada. We provide remote AI consulting for businesses in Toronto; it does not describe a local client or a staffed office.

AI consulting in Toronto · Explore the service · Compare setup options