AI Legal Trainer
TextPrompt Response Writing SFT
As an AI Legal Trainer at Pareto, I trained and evaluated large language models on complex legal scenarios. My work involved designing prompts, assessing AI responses, and developing structured outputs for benchmarking. I focused on improving model performance to meet professional legal standards.
• Designed and tested multi-step legal prompts simulating drafting, analysis, and advisory tasks
• Evaluated AI outputs for legal accuracy, reasoning quality, and adherence to real-world standards
• Identified reasoning and factual weaknesses, providing targeted feedback to enhance models
• Developed structured performance benchmarks reflecting industry best practices
As an AI Legal Trainer at Pareto, I trained and evaluated large language models on complex legal scenarios. My work involved designing prompts, assessing AI responses, and developing structured outputs for benchmarking. I focused on improving model performance to meet professional legal standards. • Designed and tested multi-step legal prompts simulating drafting, analysis, and advisory tasks • Evaluated AI outputs for legal accuracy, reasoning quality, and adherence to real-world standards • Identified reasoning and factual weaknesses, providing targeted feedback to enhance models • Developed structured performance benchmarks reflecting industry best practices
2026 - Present