Prompt Engineering & Human-in-the-Loop Safety (AI Healthcare Platform)
TextPrompt Response Writing SFT
Developed GPT-powered symptom triage and care-level routing using human-in-the-loop safety fallbacks, creating prompts and standardized flows. Designed, validated, and refined prompt strategies for symptom classification and triage tasks. Established safety escalation and decision frameworks involving AI-assisted and manual label validation.
• Created initial and iterative GPT prompt libraries for AI triage.
• Collaborated with subject matter experts for human-in-the-loop annotation.
• Generated structured clinical text data for supervised training and model evaluation.
• Developed escalation tags, performance metrics, and evaluation rubrics for triage accuracy.
Developed GPT-powered symptom triage and care-level routing using human-in-the-loop safety fallbacks, creating prompts and standardized flows. Designed, validated, and refined prompt strategies for symptom classification and triage tasks. Established safety escalation and decision frameworks involving AI-assisted and manual label validation. • Created initial and iterative GPT prompt libraries for AI triage. • Collaborated with subject matter experts for human-in-the-loop annotation. • Generated structured clinical text data for supervised training and model evaluation. • Developed escalation tags, performance metrics, and evaluation rubrics for triage accuracy.
2025 - Present