BRITE Institute is seeking computer scientists and AI engineers to help build a high-impact offline retrieval-augmented generation (RAG) system designed to support medical care in disaster zones, conflict environments, and other settings with limited or no internet access. In these environments, clinicians and first responders often operate with incomplete information under extreme time pressure—conditions where errors can be fatal. This project focuses on developing a robust, lightweight AI system capable of delivering accurate, evidence-based medical guidance entirely offline, drawing from vetted clinical resources. You will contribute to designing a system that integrates local knowledge bases, efficient retrieval pipelines, and optimized language models to function reliably on constrained hardware. This is an opportunity to directly influence patient outcomes in some of the most challenging conditions on earth.
We are looking for volunteers with strong skills in computer programming (e.g., Kotlin), AI engineering , and software development, particularly those with experience building applications, working with LLMs, or developing retrieval systems.
Here are some desired skills, ranked by priority. You do not need to have all these skills. Please let us know which skills you have:
Android/Kotlin/C++
BM25 / FTS5
medical LLM fine tuning
ONNX Runtime/ LiteRT python evaluation on android (langsmith/custom)
medical RAG / offline AI integration
Object Box or vector database experience
Android UX/UI
Above all, we are looking for individuals who are motivated to apply their technical expertise to a mission where performance, reliability, and usability are not just technical goals—but matters of life and death.