-
-
Achievements
-
Learning Tools
As a high school student, I see firsthand how stressful and expensive standardized testing and tutoring can be. Some students can afford premium private tutors to help them with the SAT or difficult math concepts, while others are stuck with static textbooks that don't adapt to how they learn. I wanted to democratize access to elite-level tutoring. The inspiration for Bright Path was to create an AI tutor that doesn't just hand you the answers, but actually acts like a world-class teacher identifying your weak spots, adapting the difficulty, and using the Socratic method to guide you to the solution yourself. What it does Bright Path is a personalized, adaptive learning dashboard powered by AI. It features four main pillars: Adaptive SAT Prep:** Diagnoses a student's weak areas in math and reading/writing, dynamically adjusting the difficulty of questions to optimize their score trajectory. Math Skills Builder: Doesn't just give practice problems; it identifies foundational gaps. For example, if a student struggles with finding the roots of a quadratic equation like $ax^2 + bx + c = 0$, the AI realizes they might need a refresher on factoring first, and scaffolds customized lessons to bridge the gap. Reading & Comprehension Tutor:** Guides students through complex passages using Socratic questioning, adjusting the reading level based on their comprehension speed.
- Personalized Study Coach: Generates weekly study plans, sends reminders, and tracks overall progress to keep students accountable.
How I built it To build this rapidly and effectively, I split the architecture into two distinct parts: the "Body" and the "Brain." The Body (Frontend):** I used an AI builder to rapidly scaffold a clean, modern React dashboard using Tailwind CSS. By letting AI handle the CSS and layout boilerplate, I saved dozens of hours of manual styling.
The Brain (Backend & Logic):** I wired the React frontend directly to the Gemini 2.0 Flash API. All the adaptive logic, Socratic questioning, and curriculum generation happen via advanced system prompts I engineered for Gemini.
// Example of the core API integration routing user queries to Gemini
async function fetchTutorResponse(studentInput, context) {
const response = await ai.generateText({
model: 'gemini-2.0-flash',
prompt: `Act as a Socratic tutor. The student asked: ${studentInput}. Context: ${context}. Do not give the direct answer. Provide a hint.`
});
return response.text;
}
Log in or sign up for Devpost to join the conversation.