RAISEF
Case Study 1: AI-Driven Healthcare Diagnostics
Design
Turn intent into testable requirements, roles, and safeguards
Specify data boundaries, choose explanation and oversight approaches, and plan security, robustness, and privacy from the start.
In Design, teams translate the problem statement into a blueprint that links system choices to RAISEF drivers. Requirements name the people served, expected outcomes, and measurable indicators. Data needs and limits are defined, including lawful basis, minimization, and retention. Designs select interpretability and explainability approaches, document human oversight points, and include guardrails for safety and security. Risks, assumptions, and trade offs are recorded with owners and evidence to collect in testing and deployment. The output is a reviewable plan that makes responsible choices concrete and auditable before development begins.
