Independent AI Evaluator & Red Teaming Specialist
OtherTextRed Teaming
I conducted adversarial prompt testing and performed comprehensive safety audits on financial large language model outputs. My responsibilities included labeling outputs as either 'Safe' or 'Critical Fail' according to 6-dimensional audit criteria. Reports were generated to differentiate lawful AI financial responses from non-compliant or risky outputs.
• Audited LLM outputs for bias, legal, factual, and regulatory compliance
• Labeled and classified text responses based on Indian financial safety guidelines
• Utilized multi-turn adversarial strategies for robust safety assessment
• Ensured model outputs aligned with institutional and legal risk boundaries.
I conducted adversarial prompt testing and performed comprehensive safety audits on financial large language model outputs. My responsibilities included labeling outputs as either 'Safe' or 'Critical Fail' according to 6-dimensional audit criteria. Reports were generated to differentiate lawful AI financial responses from non-compliant or risky outputs. • Audited LLM outputs for bias, legal, factual, and regulatory compliance • Labeled and classified text responses based on Indian financial safety guidelines • Utilized multi-turn adversarial strategies for robust safety assessment • Ensured model outputs aligned with institutional and legal risk boundaries.
2026 - Present