Efficient fine-tuning of Phi-3.5-Mini-Instruct using Unsloth and LoRA for Medical Question Answering. Demonstrates memory-efficient training, faster inference, and domain-specific adaptation to build an accurate healthcare AI assistant.