ShikshaAI – Gemini Study Planning Agent

Inspiration

Students often spend more time planning how to study than actually studying. Creating schedules, prioritizing difficult topics, and designing revision strategies manually is stressful and inefficient.

We wanted to build an intelligent assistant that thinks like a mentor — automatically analyzing a syllabus and generating a smart, personalized study strategy. With the launch of Google Gemini’s advanced reasoning capabilities, we saw an opportunity to create not just a planner, but an autonomous learning agent.

What it does

ShikshaAI is a Gemini-powered study planning agent that:

Analyzes entire syllabi or notes
Breaks content into structured topics
Estimates difficulty and study time
Generates personalized day-wise study plans
Creates adaptive quizzes for revision
Tracks learning progress visually

Instead of a simple chatbot, it behaves like a multi-step AI system that reasons, plans, and assists throughout the learning process.

How we built it

We built a full-stack web application using:

React + Vite (frontend dashboard)
Node.js + Express (backend APIs)
PostgreSQL (data storage)
Google Gemini API for reasoning and content generation

We designed a service layer that orchestrates multiple Gemini calls:

Syllabus analysis
Topic extraction + difficulty estimation
Study plan generation
Quiz generation

This multi-step architecture allows the app to behave like an AI agent rather than a single prompt wrapper.

Challenges we ran into

Structuring large syllabus text for accurate AI reasoning
Designing prompts that consistently return structured outputs
Handling fast AI responses while still showing clear “thinking” states
Balancing clean UI with powerful functionality

We iterated heavily on prompt engineering and UX to make the system both reliable and intuitive.

Accomplishments that we're proud of

Built a complete production-style web app, not just a demo
Implemented multi-step Gemini orchestration
Created a clean, professional dashboard experience
Delivered real-time AI planning + quiz generation
Made AI usage transparent and explainable to users

What we learned

Prompt engineering is as important as code
Multi-step AI workflows feel far more intelligent than single calls
Good UX is critical for showcasing AI systems
Gemini’s reasoning capabilities enable complex educational tools with minimal infrastructure