Skip to content

THORzero9/aura-agentic-app

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

27 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Aura - Your Personal AI Learning Guide

Aura-App-Logo
Aura is an intelligent, full-stack mobile application designed to combat information overload and create personalized, high-quality learning paths for any topic.

The Problem

In today's world, learning a new skill is overwhelming. You are faced with a sea of random YouTube videos, blog posts, and expensive courses, with no clear path from beginner to expert. It's difficult to know where to start and what to learn next, leading to frustration and abandoned goals.

The Solution

Aura acts as your personal AI mentor. Simply tell the app what you want to learn, and our advanced, multi-step AI agent gets to work. It deconstructs the topic, searches for the best free resources on the web, and curates them into a logical, week-by-week curriculum tailored just for you. With Aura, you get a clear, structured path to mastery.


🚀 Key Features

  • Intelligent Path Generation: Our AI agent understands your learning goals and creates a structured, easy-to-follow plan.
  • AI-Powered Curation: We don't just search; our agent analyzes and selects the best, most relevant resources from around the web, using a two-step "Search and Curate" workflow.
  • Secure User Authentication: Full email/password and Google Sign-In support, powered by Appwrite. Includes a robust OTP email verification flow.
  • Save & View Plans: Users can save their favorite generated plans to their personal account and view them anytime.
  • Polished Native Experience: A beautiful, responsive, and user-friendly native Android application built with the latest Jetpack Compose standards.
  • Web Frontend: A fully functional web app built with React, deployed on Netlify, for universal accessibility.

🛠️ Tech Stack & Architecture

Aura is a full-stack application built with a modern, decoupled architecture.

Frontend (Android)

  • Language: Kotlin
  • UI: Jetpack Compose
  • Architecture: MVVM (Model-View-ViewModel) with a Repository Pattern
  • Networking: Retrofit
  • Dependencies: Lottie (for animations), Jetpack Navigation

Frontend (Web)

  • Framework: React
  • Deployment: Netlify

Backend (The Agent)

  • Framework: Python with FastAPI
  • Deployment: Google Cloud Run for scalability and management.
  • AI Orchestration:
    • Google Generative AI (Gemini 2.5 Flash): Used for all reasoning tasks, including deconstructing topics and curating resources.
    • Tavily Search API : Used for all web searches to find high-quality, relevant learning materials.

Backend-as-a-Service (BaaS)

  • Appwrite :
    • Appwrite Auth: Manages all user accounts, including Email/Password, Google Sign-In, and OTP verification.
    • Appwrite Databases: Provides the persistent cloud database for storing all user-saved learning plans.

🧠 How It Works: The AI Agent

The core of Aura is its intelligent agent, which follows a four-step process to create a high-quality learning plan:

  1. Deconstruction: The user's topic is sent to a Generative AI model (Gemini) to be broken down into four logical, weekly sub-topics.
  2. Search: For each sub-topic, the Tavily search API is used to find relevant, high-quality learning resources like articles and videos.
  3. Self-Correction: An AI-powered sanity check evaluates the initial search results. If they aren't relevant for a beginner, the agent refines and re-runs the search query.
  4. Curation: The final, validated search results are passed to the AI model one last time to select the top 2-3 resources for each week and format them into a clean JSON response.

This agentic approach ensures the generated plans are more relevant and higher quality than a single, simple API call.

⚙️ Setup and Installation

Prerequisites

  • Android Studio (latest version)
  • Python 3.10+
  • An active .env file in the aura-backend directory with valid API keys for:
    • GOOGLE_API_KEY
    • TAVILY_API_KEY
    • APPWRITE_PROJECT_ID
    • APPWRITE_API_KEY
    • APPWRITE_DATABASE_ID

Running the App

  1. Backend:

    cd aura-backend
    python3 -m venv venv
    source venv/bin/activate
    pip install -r requirements.txt
    python3 main.py
  2. Android App:

    • Open the aura-android folder in Android Studio.
    • Update the BASE_URL in ApiClient.kt to point to your local machine's IP address.
    • Build and run the app on a physical device or emulator.

🎬 Demo Video

https://youtu.be/qxn4Sp8i0as?si=Xiy0qMDLqaH4YPQ6

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors