An AI-powered 4-year academic planner built for Berea College students.
Topic: Solve a Campus Problem
Planning four years of college at Berea is genuinely difficult. Students must satisfy GEM requirements (Ways of Knowing, Richnesses, the Learning & Inquiry sequence), fulfill major prerequisites in the right order, stay within Berea's 4-credit-per-semester structure, and keep their career goals in mind — all at the same time. Most students do this manually with a spreadsheet or rely on sporadic advisor check-ins. Mistakes are easy to make and costly to fix.
Lumen generates a complete, constraint-satisfying 8-semester course plan in under a minute, then layers on personalized AI career advice.
User journey:
- A student opens the planner and completes a 4-step form: major(s), academic interests, hobbies, and career goals.
- Lumen runs a constraint-based scheduling algorithm and produces a full 4-year plan with real courses, GEM categories, and prerequisites correctly ordered.
- The plan page shows each semester's courses grouped by year, with a stats summary (total credits, GEM coverage, etc.).
- An AI career advisor (powered by Claude) analyzes the plan and provides specific advice: which courses to prioritize, skills to develop, companies to target, Berea alumni to find on LinkedIn, and concrete ways to gain experience.
- Frontend: Next.js 16 (App Router), React 19, TypeScript, Tailwind CSS
- Components: shadcn/ui, Radix UI, Lucide React
- AI: Vercel AI SDK, Anthropic Claude Sonnet 4
- Hosting: Vercel, Vercel Analytics
- Tooling: pnpm, v0 (AI-assisted UI scaffolding)
The scheduling algorithm was the hardest part. Satisfying all constraints simultaneously — prerequisite chains that span multiple years, GEM category coverage, credit limits per semester, major course caps, and the L&I 100 to 400 sequence gated by class year — required carefully ordering each placement pass. If prerequisites were resolved in the wrong order, the whole plan would break. We also had to build fallback placeholder logic for cases where a specific course could not be scheduled in a valid slot. Getting the output to feel intentional rather than just technically valid took many iterations.
The AI integration also required encoding Berea-specific institutional context directly into the system prompt — the labor program, the GEM framework, the alumni network — to produce advice that was actually useful rather than generic.
- The scheduler reliably produces valid 4-year plans across multiple major combinations without any manual tuning.
- The AI career advice feels specific to Berea students because the prompt is built around Berea's actual culture, programs, and community.
- The full flow from onboarding to a complete plan with career guidance takes under a minute.
Live app: https://v0-lumen-berea.vercel.app/
No installation needed. Open the link and get started.
| Name | Role |
|---|---|
| Essey Araya | Backend Lead |
| Bright Feitsop | Data Lead |
| Naod Ksmu | Frontend Lead |