A use case defines how a person or organization applies a system to accomplish a goal. For AI products, it should address a real customer need and describe a task where AI can create measurable impact through augmentation or automation. section 2.0 section 10.4
Shape a use case by checking:
- Whether the task reflects an actual user need rather than an attempt to add AI for its own sake. section 2.0
- What business impact it could produce, how technically feasible it is, and whether it fits company strategy, goals, and skills. section 2.0
- How users will respond to probabilistic outputs, failures, uncertainty, and opportunities to modify or reject results. section 10.4
- How evaluation metrics, real-world performance data, edge cases, and user feedback will reveal whether the use case works. section 8.3 section 10.4
A chatbot without a job
A chatbot built without a clear use case can frustrate users instead of improving service. A stronger definition identifies the user task, the intended value, and the measurable effect before selecting an AI approach. section 2.0
Use the opportunity-discovery process to move from sourcing possibilities to prioritizing and shaping the most valuable AI use cases. section 2.0
