Axis

Inspiration -Course materials live across PDFs, slides, and links. Students waste time organizing; professors repeat formatting. We set out to build a standardized, AI-native course that’s ready on day one.

What it does -Axis converts professor uploads into rich, consistent course pages—units, lessons, key concepts, practice, and assessments—and a student dashboard showing current topics, deadlines, and recommended review. Uploads provide the outline; trusted web sources fill gaps with citations.

How we built it -Next.js + TypeScript + Tailwind + shadcn/ui for the UI -Supabase (Postgres, Storage, Auth/RLS) for users, roles, courses, and files -Gemini to structure/expand content; Bing Web Search API to augment with vetted sources -YAML/MDX course schema rendered into lesson pages -Professor UI with drag-and-drop upload, AI preview, and publish

Accomplishments we’re proud of -A repeatable course template for consistent classes -End-to-end ingestion: upload → outline → augmented lessons → approval → live -A focused student dashboard that surfaces what matters now

What we learned -Pipeline works best as steps: outline → expand → cite → QA -Citation checks and overlap scoring improve reliability -Simple, clear UI increases adoption

What’s next for Axis -In-lesson practice generators and auto-graded quizzes -Analytics for professors (engagement, confusion hotspots) -Duke pilot (free) to refine, then license to additional universities

Built With

• next.js
• react
• shadcn/ui
• tailwind-css
• typescript