A manual override lets users adjust an automated forecast when it conflicts with local knowledge or real-world conditions. It gives them control while preserving feedback for later system improvement. section 12.3.2
A practical override workflow includes:
- Spot errors by comparing predictions with historical patterns and local knowledge. section 12.3.2
- Address errors by making a manual adjustment when the forecast does not match user insight. section 12.3.2
- Report errors through mechanisms such as Flag for Review so feedback can inform later updates. section 12.3.2
A planner corrects a forecast
A planner notices that a forecast diverges from local conditions, adjusts it manually, and flags the issue for review. The feedback is considered during a later fine-tuning round. section 12.3.2
