A generative-AI-assisted platform for rapidly creating interactive,accurate 3D explainers for all of biomedical science and medicine (TABA Supplement)

Project Summary/Abstract Our currently active NIH Phase I SBIR project focuses on the development of a human-vetted, AI-augmented web platform that transforms complex scientific text into accurate, pedagogically sound visuals. With the rise of large language models (LLMs) from providers like OpenAI, DeepSeek, and Anthropic, information access has dramatically improved, but these tools remain text-dominant. Our product addresses the needs of the large population of visual learners in STEM by generating curated graphics that enhance comprehension and retention. To maintain a first-mover advantage, we now seek to supplement our core R&D effort with critical business and technical intelligence services that will guide us through an increasingly complex competitive and regulatory landscape—particularly as AI visualization tools become more relevant to healthcare and medical education. This Technical and Business Assistance (TABA) supplement will fund third-party expert services and resources that are outside the technical scope of our current Phase I award, but are essential for de-risking commercialization. These activities include benchmarking emerging LLM and VLM models across multiple providers (e.g., OpenAI, Google, DeepSeek) to identify cost-effective, high-performing platforms for integration, and engaging HIPAA compliance and FDA regulatory consultants to ensure future readiness for healthcare deployment. Since our initial application, the FDA has released new draft guidance on AI-enabled software functions (FDA-2024-D-4488), signaling regulatory scrutiny of tools like ours when deployed in clinical workflows. The proposed TABA work will directly accelerate the impact and reach of our Phase I innovation by informing our model selection, compliance approach, and product roadmap. These activities will also position us to enter healthcare and health education markets more responsibly and competitively. By leveraging third-party assessments and regulatory strategy experts, we aim to ensure that our AI-enabled visualization system is not only technically robust but also aligned with evolving standards in safety, privacy, and market expectations.