AI/ML Engineer Consultant – LLM Fine-tuning and AI Model Optimization
Led the fine-tuning of large language models (LLMs) on domain-specific text data across healthcare and agriculture. Integrated and benchmarked open-source LLMs for deployment in clinical and field environments, optimizing for both performance and efficiency. Managed the end-to-end model lifecycle, including experiment tracking and scalable hosting on AWS SageMaker. • Fine-tuned ClinicalBERT and LLaMA models on clinical and agronomic text datasets for specialized diagnostic and Q&A applications. • Applied 4-bit quantization with bitsandbytes to optimize LLM inference for memory footprint and deployment in resource-constrained settings. • Evaluated and prepared datasets to enable RAG and context-aware information retrieval from highly specialized knowledge bases. • Established model observability and compliance for HIPAA-aligned healthcare use cases using Grafana and Dockerized services.