Health disparities severely limit access to preventative care in underserved regions, such as remote Greek island communities. Addressing this critical gap, "HealthBot" was conceived - an innovative, AI-powered health kiosk developed by motivated high school students to provide immediate, accessible initial health insights and enhance health equity. This paper details HealthBot's design, which uniquely integrates diverse sensors (weight, height, custom MAX86150 biometrics) with AI for non-invasive preliminary assessments, including cloud-based ML-driven skin lesion analysis and demographic estimation, leveraging a Large Language Model (GPT-4 mini) to interpret data and offer personalized, non-diagnostic wellness advice. Initial usability feedback gathered from 24 participants indicated manageable workload, although perceived system performance was rated lower due to sensor calibration issues during the prototype testing phase. In conclusion, while further validation is required, the HealthBot project demonstrates a feasible model for enhancing health equity through readily available, technologically advanced community health tools, serving as a compelling case study for project-based STEM education where student innovators address real-world problems.