GenAI Engineer Project - Denys

GenAI Engineer Project - Denys

icon
This track equips learners to proficiently create, integrate, test, and deploy AI-driven applications and prompt engineering solutions, leveraging modern AI models and real-life cloud environments. This is a back-end oriented project.
icon
Project Definition

To develop a web application for learning about space using AI to generate questions and 3D objects in the scene with three.js. The application is aimed at children and anyone interested in space, offering an interactive learning experience.

Description of the application:

  • Users learn about space and answer questions using 3D objects in the scene.
  • The application will include user registration, login, question generation, answer processing, and feedback.
  • Plans include adding AR functionality for mobile devices.

2. Functional Requirements

2.1 Registration and Login

  • User registration via email.
  • Confirmation email after registration.
  • Login via email and password.
  • Password recovery.

2.2 Tutorial for New Users

  • A short tutorial is shown upon the first launch, explaining the interface and functionality of the app.
  • Option to skip or go through the tutorial.

2.3 Question Generation and Displaying Objects in 3D Scene

  • AI-generated questions.
  • 3D objects related to the question (e.g., planet or satellite) are displayed in the scene.
  • Interaction with the object (rotation, zooming).

2.4 Answer Processing

  • Correct answer: Praise and detailed explanation after selecting the correct answer.
  • Incorrect answer: Explanation of the correct answer after an incorrect response.

2.5 End of Learning and Continuation

  • After each question, the app asks if the user wants to continue learning.
  • If continued, a new question is generated based on the user’s knowledge.

2.6 Log Out

  • Option to log out from the app via the settings menu.
icon
Project Links
icon
Project Resources
  • Add any resources about the project
Resources

3. UI Requirements

  1. Main Screen
    • Includes a 3D scene, answer selection panel, and a button to continue.
  2. 3D Scene
    • Objects related to the questions are displayed (e.g., planets).
    • Ability to interact with objects.
  3. Feedback
    • After answering, feedback is provided with detailed explanations.
  4. Navigation Buttons
    • Buttons for navigating between screens: "Next", "Back", "Finish", "Continue".

4. Architecture and Technologies

Backend

  • Programming Language: Python
    • Flask for MVP.
    • Django for future upgrades.
  • Database:
    • MongoDB or PostgreSQL for storing data.
  • AI:
    • Groq
    • Using GPT-3 or Olama for question generation.
    • AI for generating textures and objects for 3D scene.

Frontend

  • Programming Language: JavaScript (ES6+).
  • Framework: Vue.js or React.

Library

Three.js for 3D scenes.

5. Implementation Stages

5.1 Preliminary Stage

  1. Define the database structure.
  2. Design the app architecture.

5.2 Backend Development

  1. Create API.
  2. Integrate AI.
  3. Set up the database.

5.3 Testing

  1. API and database testing.
  2. AI integration testing.

5.4 Deployment

  1. Host the app on the Hostinger server. ??
  2. Set up the database and security.
icon
MVP
icon
Timeline
Week
Topic
Project Tasks
Learning (Read/Watch/Exercise)
Completed
1
creating docs
yes
2
Basic app
creating an app with basic modules (api gateway, signIn login)
yes
3
Creating service with generating questions
Conecting an app to Crog and make an service that will generate questions
yes
4
creating frontend
create a frontend that fits the requirements of MVP
yes
5
creating auth service
create authentification logic
yes
5
creating question history service
Creating logic of storing questions
yes
6
creating connections between services (main logic)
combine services together
yes
7
fixing bugs
testing debuging
yes
8
finalising api tests
testing debuging
yes
8+
icon
MVP requirements
  • Registration and Login
  • Tutorial for New Users
  • Question Generation
    • AI-generated questions.
  • Answer Processing
    • Correct answer: Praise and detailed explanation after selecting the correct answer.
    • Incorrect answer: Explanation of the correct answer after an incorrect response.
icon
V2 optional requirements
  • Add models generator
  • Add 3D features in frontend app
  • Add an advanced tutorial with a 3d scene for an app
  • testing