Prompt Engineer & LLM Adversarial Tester (NairaSmsPulse)
TextRed Teaming
I engineered complex prompt structures and conducted adversarial testing of Google Gemini to sanitize and evaluate unstructured financial text data. My work ensured the identification of model vulnerabilities and edge-case failures, improving the system's output integrity. This experience involved systematic documentation and reporting of ambiguous data actions for better categorization and model robustness.
• Adversarial prompt engineering and red-teaming against LLMs.
• Categorization and documentation of ambiguous edge cases where LLM failed.
• Evaluation of sanitized model outputs for compliance and quality.
• Improved data pipeline resilience during AI and connectivity failures.
I engineered complex prompt structures and conducted adversarial testing of Google Gemini to sanitize and evaluate unstructured financial text data. My work ensured the identification of model vulnerabilities and edge-case failures, improving the system's output integrity. This experience involved systematic documentation and reporting of ambiguous data actions for better categorization and model robustness. • Adversarial prompt engineering and red-teaming against LLMs. • Categorization and documentation of ambiguous edge cases where LLM failed. • Evaluation of sanitized model outputs for compliance and quality. • Improved data pipeline resilience during AI and connectivity failures.
2026 - Present