R. David Parker

Research Interests

Artificial intelligence, data, missed opportunities

Retrieval-Augmented Generation (RAG), LLMs, and Data Ecosystem Modernization

My most recent work expands these approaches by developing retrieval-augmented generation (RAG) systems and LLM-driven analytic architectures capable of integrating decades of complex, heterogeneous data into unified research ecosystems. I designed and developed, deployed a local, security constrained RAG platform that ingests 40 years of data PDFs, XML, spreadsheets, relational databases, and unstructured archival content into a high-performance retrieval layer that supports faculty research, operational analytics, and compliance. This system processes large collections of unstructured research records without exposing sensitive data to external environments, offering a scalable model for clinical informatics, biomedical research, and health-system data modernization. This RAG framework directly informs my current scholarly trajectory: developing domain-specific LLMs and multi-modal retrieval systems tailored for educational and research text, biomedical literature, operational oceanographic, environmental sciences, and research administration workflows. I view RAG (and similar tools) enhanced architectures as a critical bridge between data system complexity and actionable intelligence supporting decision making, accelerating translational research, and improving data governance across institutions.

AI, Machine Learning, and Computational Analytics

A core throughline in my work is the development and application of machine learning and natural language processing to clinical and population level problems. I use computational methods such as classification models, regularization techniques (lasso, ridge), dimensionality reduction, predictive modeling, and advanced regression architectures to identify missed diagnoses, forecast risk, and improve care quality using large claims, EHR, and surveillance datasets. My work integrates both interpretable models and high-performance predictive pipelines, balancing methodological rigor with clinical practicality. Several of my projects in Alaska, West Virginia, and South Carolina demonstrate how NLP can be deployed to extract latent information from provider notes, case narratives, or disparate record systems to detect patterns that structured data alone cannot capture. These efforts have been embedded into multiple domain workflows to identify movement of troop and other defense resources, high-acuity patients, understand drivers of chronic illness, and improve addiction, behavioral health, and infectious-disease outcomes.

Education

• PhD, epidemiology & biostatistics, University of South Carolina

Professional Experience

Senior leader in higher education with a 25+ year record of advancing research strategy, academic innovation, and cross-sector collaboration at major public universities and academic medical centers. Background includes over $25M in funded research, leading initiatives in public health, AI integration, healthcare innovation, and research infrastructure, and 40+ peer-reviewed publications & 100+ technical and professional presentations. Vision-driven, data-informed leadership style that strengthens institutional capacity, supports faculty scholarship, and aligns research growth with strategic mission.

Other Links

{{https://orcid.org/0000-0002-1955-7445,ORCID}}

CV