College of Dentistry Leverages Generative AI Using Amazon Q Business for Point Of Care Integration of Curriculum Knowledge

"Bridging Theory and Practice: Integrating Knowledge Across the Dental Education Continuum"

The Ohio State University College of Dentistry delivers a comprehensive, multi-year curriculum that spans over one hundred courses across the basic and clinical sciences. Organized into 11 academic divisions, the pre-doctoral Doctor of Dental Surgery (DDS) and Dental Hygiene (DH) programs follow a structured, sequential progression—four years for DDS and two years for DH. Instruction is designed to build cumulatively, with students advancing through didactic and clinical coursework in a relatively linear fashion. Throughout their training, students are expected to recall, synthesize and integrate knowledge from thousands of hours of lectures and clinical protocols, applying this information to make informed, patient-centered clinical decisions.

Addressing these challenges necessitates innovative solutions that bridge the gap between information recall and clinical application. A critical need is the timely integration of knowledge across the basic and clinical sciences, enabling seamless access to comprehensive, contextually relevant information throughout the curriculum. Such integration is essential for supporting optimal patient care and promoting academic success.

The Ohio State University College of Dentistry implemented the use of Amazon Q Business, a generative AI–powered assistant that can answer your questions, generate content, create summaries, and can help quickly describe complex workflows, to create an intelligent knowledge management system that integrates the entire dental curriculum into a searchable, AI-powered platform–a Point Of Care Academic Assistant (POCAA)–to deliver information. This solution incorporates generative AI capabilities that allow students and faculty to query course materials, clinical guidelines and educational resources using natural language, receiving contextual, evidenced-based answers that span and integrate multiple areas of dental knowledge. Generative AI helps students and faculty quickly identify the most relevant information sources. The source-attributed responses are specifically drawn from the college’s own curriculum. The latter is important because information is not derived from any external sources. In short, these responses support informed decision-making in both academic and clinical contexts.

“Amazon Q Business is a powerful tool that enables us to enhance our curriculum by providing immediate access to integrated information for both students and faculty,” said Dr. Carroll Ann Trotman, dean of the College of Dentistry. “Our implementation—developed in close collaboration with the AWS team—represents a strategic approach to optimizing educational outcomes.”

"Behind the Scenes: How Amazon Q Business Powers Personalized Learning"

The implementation of Amazon Q Business for Ohio State's dental curriculum employs sophisticated RAG-based architecture that begins with comprehensive document processing capabilities handling multiple formats including PDFs, PowerPoint presentations, lecture transcripts, and course metadata through parsing algorithms that normalize and segment content into meaningful chunks while maintaining structural integrity and contextual relationships. The system utilizes a vector-based retrieval system with dense passage retrieval (DPR) where document segments are encoded into high-dimensional vectors for efficient matching between natural language queries and relevant content. When students submit queries, the system processes them using large language models (LLMs) for intent interpretation, performs semantic search operations across the knowledge index, and generates responses incorporating multiple ranking factors including semantic relevance, document authority, content freshness, and source reliability, while implementing retrieval-augmented generation to ground responses in source materials with comprehensive citation tracking. 

A distinguishing feature is the tailored feedback mechanism centered on an intuitive thumbs up or down interface that allows immediate user feedback on response quality, supplemented by optional text fields for detailed explanations. Feedback data undergoes regular aggregation and analysis to identify patterns in response effectiveness, common confusion areas, and improvement opportunities. This creates a dynamic learning loop where the system continuously adjusts relevance rankings, refines response generation parameters, and optimizes content delivery based on actual user interaction data.

Making an Impact :"Reducing traditional searching inefficiencies through 100% reliable sources"

The technical architecture directly translates into tangible benefits for dental students. The system provides instant access to curriculum-wide knowledge, enabling contextual understanding across courses and topics through natural language interaction. Students receive consistent and accurate responses backed by official course content, available continuously for study support and concept clarification. “Giving the sources of the answers is so important, as we can be confident that the answers are correct and valid,” said one dental student piloting POCAA. The feedback system ensures that response quality improves over time, creating an increasingly effective learning resource.

Amazon Q Business leverages the college's extensive repository of course materials, including lecture notes, clinical case studies, procedural guidelines, research papers and assessment materials spanning basic sciences through advanced clinical specialties through the most recent course content. By indexing content across all dental disciplines and analyzing query patterns and user feedback, AWS Q Business continuously optimizes its responses to promote accurate, comprehensive answers that integrate multiple areas of dental knowledge. Using this curriculum-aware, adaptive approach built on AWS, students reduce information we hope to significantly reduce search time by students compared to traditional manual methods. 

One student piloting POCAA  said, “When I used the AI tool to research about local anesthetic dosing, it gave me a clear and concise answer and then followed up to teach and remind me of the formula.” By combining a large language model with a specified knowledge source into a single AI model, POCAA can deliver clear answers while maintaining the importance of clinical education. 

"Sustainable Innovation: Managing 9,000+ Artifacts Across 100+ Courses"

The Ohio State College of Dentistry has incorporated cutting-edge natural language processing and machine-learning capabilities with automated content indexing to eliminate manual searching through disparate educational resources. The fully managed Amazon Q Business can process and respond to queries 24/7 with minimal administrative oversight, automatically updating its knowledge base as new course materials and clinical guidelines are added. “By building a platform that can integrate with key data sources – such as our learning management system, lecture capture software and SharePoint we are able to streamline the new data ingestion and historical data retention of this tool. With over 9,000 total artifacts ingested over 100+ courses, sustainability was a critical feature when implementing Amazon Q Business.” says Michael Douglas, assistant dean for Information Technology at the College of Dentistry.

With Amazon Q Business providing end-to-end security out of the box, Ohio State can concentrate on productivity gains like reducing the research and preparation timelines that traditionally required students to manually search through multiple textbooks, course notes and clinical references used across dental education. The system automatically processes new content uploads, ensuring that the most current information is always accessible.

Amazon Q Business emerged as the optimal solution for our use case by enabling us to meet all project goals within a single, integrated platform. Its ability to connect directly to Canvas via API interfaces was essential, allowing seamless ingestion of the majority of our source content. Key selection criteria included rapid deployment capabilities, robust identity and access management, and advanced output guardrails to ensure response accuracy and role-based output. Collaborating closely with the AWS Account Team provided invaluable access to subject matter experts, enabling us to co-develop, refine, and deploy POCAA with precision and speed. This partnership was instrumental in transforming our curriculum into a dynamic, AI-powered knowledge system that supports both sustainable innovation and scalable impact.

The POCAA application is being rolled out to college faculty through a pilot phase as we refine the application responses, develop query guardrails, and evaluate a sourcing accuracy model to ensure this application sources the most appropriate content. As the college expands integration across additional educational platforms, we expect further learning improvements. Student interaction data and query patterns help refine the AI's understanding of educational needs and improve response accuracy, creating a continuous learning cycle that benefits both current and future students. We plan to evaluate educational outcomes, with a particular focus on the time faculty and students spend searching for curricular content and navigating complex clinical scenarios. Moreover, the platform is highly scalable—it can be expanded to integrate curricular content across specific colleges, the entire university, or narrowed in scope to support individual courses within a single college.