This project aims to develop safety-critical, human-centered AI systems across various industries, including healthcare, transportation, and automotive. It provides a structured, multi-faceted approach to the design and evaluation of AI and large foundation model systems, ensuring they are safe, reliable, and inclusive for three high-stakes applications. In healthcare, the project leverages AI-driven telemonitoring to improve patient care by focusing on early identification of high-risk individuals. Specifically, it introduces predictive models to monitor and flag patients who may be at elevated risk, offering caregivers actionable insights through an intuitive LLM interface. By translating complex predictions into comprehensible information, the project aims to support caregivers in delivering cost-efficient, personalized home care that adapts to each patient’s needs and helps prevent adverse events. In transportation, the project emphasizes safety and inclusivity in transportation, particularly for vulnerable road users who may be overlooked by conventional AI systems. By collecting diverse, high-quality data that represents a range of user behaviors and mobility needs, the project builds a robust benchmark for detecting and tracking diverse groups of road users. This data set enables the refinement of detection algorithms to better recognize and respond to the specific needs of users with reduced mobility. Additionally, an inclusivity index will be established to assess how well AI-driven digital models of transportation systems represent diverse user groups. This tool will serve as a resource for stakeholders in urban planning and digital system development, helping them understand and enhance the inclusivity of their solutions. In the automotive domain, the project targets productivity enhancements by adapting generalist language models to domain-specific tasks. By tailoring these LLMs to process and generate insights from the large volume of text-based artifacts in engineering, this component reduces the time engineers spend on documentation, freeing them to focus on design and innovation. This approach also sets a foundation for future development of specialized LLMs in other fields, such as healthcare and telecommunications, with adaptable, reusable methods and logic. Overall, this project advances the role of AI as a safe and reliable tool, laying the groundwork for a more inclusive, responsive, and effective future across different industries. This project will only use public datasets for its development and validation.