Prompt Engineering and LLM Fine-Tuning for Q&A Accuracy
TextFine Tuning
I performed prompt engineering and LLM fine-tuning to improve Q&A accuracy for domain-specific slide content on SlideCoach's AI-driven tutoring platform. My work centered on preparing and curating textual datasets for model training and validation in real-world education contexts. I used AWS Bedrock and SageMaker as the main AI infrastructure throughout the project.
• Designed and applied prompts for LLM performance improvement.
• Selected, organized, and augmented text-based datasets for supervised fine-tuning and evaluation.
• Monitored and evaluated model outputs to enhance answer relevance and correctness.
• Used internal/proprietary tooling integrated with AWS services for data ingestion and model feedback cycles.
I performed prompt engineering and LLM fine-tuning to improve Q&A accuracy for domain-specific slide content on SlideCoach's AI-driven tutoring platform. My work centered on preparing and curating textual datasets for model training and validation in real-world education contexts. I used AWS Bedrock and SageMaker as the main AI infrastructure throughout the project. • Designed and applied prompts for LLM performance improvement. • Selected, organized, and augmented text-based datasets for supervised fine-tuning and evaluation. • Monitored and evaluated model outputs to enhance answer relevance and correctness. • Used internal/proprietary tooling integrated with AWS services for data ingestion and model feedback cycles.
2025 - Present