- Assessed within a broader engineering context, i.e. ML is one of the many systems/tools of a company/organization.
- ML replaces existing systems, simplifying operations and/or maintenance.
- ML improves specific operations’ performance via (mostly accurate) predictions
#Examples
- Replacing difficult to build rule-based systems with difficult to maintain exceptions.
- Adding filtering capabilities to a customer support system.
- Note: small error rate is not critical for both scenarios.