Elimu Sense
Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

Elimu Sense is an AI accessibility platform that helps blind and low-vision learners understand visual academic content independently. This help unlock high paying visual heavy disciplines. I am visually impaired, and I built Elimu Sense because a screen reader can read the words around a diagram without giving me the diagram itself. In school and university, maps, charts, graphs, scientific illustrations, music scores and mathematical figures are often treated as optional visuals. For a visually impaired student, they are frequently the difference between understanding a lesson and being excluded from it. A learner uploads an academic image, chart, map, diagram or music score. Elimu Sense analyses its structure and converts it into an organized audio explanation. Instead of producing a vague caption such as “this is a graph,” it explains the visual in four useful layers: the overall layout, the important elements and their positions, the relationships or patterns between them, and exam-ready takeaways. The learner can then ask follow-up questions by typing or speaking and receive answers about the image. The goal is not to replace a teacher, a tactile graphic or human verification. The goal is to give a learner a fast first layer of access so they can study, revise and participate without waiting days for someone to manually describe every visual. The current scope is academic visual content for high school, university and professional learning. Elimu Sense is not designed to make medical, legal or safety-critical decisions, and it does not present uncertain interpretations as facts. For assessment-linked material or low-confidence outputs, the system is designed to flag content for human review. Learners can also report an incorrect description. Technically, the system combines image understanding, structured extraction, academic context and audio delivery. A simple computer-vision or rules-based system can detect edges, colours or text, but it cannot reliably explain that a contour line represents increasing elevation, that two variables are correlated, or that a labelled anatomical structure has a specific relationship to another structure. Elimu Sense uses a multimodal model to reason about those relationships and then formats the result into a predictable, accessible learning experience. This is a “Build for Your World” project in the most literal sense. I am building the tool I needed when visual course material quietly closed entire subjects to me. Elimu Sense is intended to make Geography, STEM, Music and other visual disciplines available on equal terms, beginning with the learners who are usually expected to adapt to inaccessible systems.