Bernadette Lorden
AI Safety Evaluator | Risk & Policy Analysis
New to Contra
Bernadette is ready for their next project!
AI evaluation case study focused on assessing model responses within healthcare workflows and Epic EHR documentation scenarios.
Demonstrates ability to identify clinical risks, workflow misalignment, and patient safety concerns while applying structured evaluation and policy-based reasoning.
Combines AI evaluation expertise with real-world healthcare system knowledge to ensure safe, accurate, and compliant outputs.
0
5
AI safety evaluation case study focused on identifying harmful bias and stereotype reinforcement in model responses.
Demonstrates structured analysis using policy-based reasoning, risk classification, and severity scoring to assess fairness and safety alignment.
Designed to reflect real-world AI evaluation workflows requiring strong judgment and consistency.
0
8
Red team AI evaluation case study demonstrating adversarial prompt design and analysis of model responses for manipulation and exploitation risks.
This example highlights my approach to identifying policy gaps, assessing model behavior under edge-case scenarios, and applying structured risk classification and severity scoring.
Focused on detecting subtle failure modes and producing clear, defensible justifications aligned with Trust & Safety standards.
0
9
:Example of how I evaluate AI-generated responses using structured Trust & Safety frameworks, including risk classification, severity scoring, and policy-based reasoning.
This case study demonstrates my ability to identify nuanced safety gaps, assess model behavior, and produce clear, consistent evaluation outputs at scale.
Open to AI evaluation, red teaming, and safety-focused contract work.
0
12