LUMINA

• 

  User enters learning goal, skill level, available time, and deadline.

• 

  LUMINA generates a structured, step-by-step learning plan using Gemini reasoning.

• 

  Clear weekly breakdown with phases and tasks designed for real-world learning.

Inspiration

Many learners know what they want to learn, but get stuck on how to plan it. We noticed that students often waste time jumping between tutorials, courses, and platforms without a clear roadmap. This leads to confusion, burnout, and inconsistent progress—especially when preparing for placements, exams, or skill transitions.

We wanted to solve a simple but real problem: “Given a learning goal and limited time, what should I study each day?” That question became the inspiration for LUMINA.

What it does

LUMINA turns any learning goal into a clear, step-by-step study plan. The user enters: Their learning goal (e.g., Python for placements, Data Science basics, Music theory) Skill level Available time per day Deadline

LUMINA then generates a practical, weekly learning planner with: Phases and milestones Daily task breakdowns Clear priorities (what to focus on first) Interview/exam-oriented structure where applicable The output is designed to look like a real planner a student would follow, not a long technical report

How we built it

We built LUMINA using Google AI Studio with Gemini 3 Pro as the core reasoning engine. Gemini handles goal understanding, feasibility reasoning, and plan generation Carefully designed prompts guide the model to produce: Task-based planners Weekly milestones Practical, time-bounded outputs A lightweight frontend (HTML, CSS, JavaScript) presents the planner in a clean, readable format The focus was on reasoning quality and usefulness, not just text generation.

Challenges we ran into

One major challenge was avoiding over-structured, report-style outputs. Early versions looked too academic and not user-friendly. We solved this by: Iterating on prompts Shifting from “sections and analysis” to checklists and weekly planners Constantly asking: “Would a real student actually follow this?” Balancing detail with simplicity was another key challenge.

Accomplishments that we're proud of

Built a goal-to-plan system, not just a chatbot Achieved clean, practical planner outputs with minimal UI Designed a solution that works across multiple domains, not only coding Successfully used Gemini’s reasoning abilities to generate structured learning paths.

What we learned

Prompt design is as important as the model itself Users value clarity and actionability more than complex explanations AI is most powerful when it guides decisions, not just answers questions Good UX sometimes means less information, not more

What's next for LUMINA

Next, we plan to: Add confidence scores and adjustment suggestions if time or skill level changes Allow users to regenerate plans week-by-week Export planners as PDF or calendar-ready formats Support progress tracking and reminders LUMINA aims to become a personal learning planner, not just a one-time generator.

Built With

• css
• gemini3api
• googleaistudio
• html
• javascript