QuGenAI

QuGenAI is an AI-powered exam content generator designed to help educators, certification providers, and trainers create high-quality, data-driven assessments in minutes. Built for flexibility and scalability, QuGenAI streamlines exam development workflows, ensuring efficiency, accuracy, and reliability.

Target Audience

• Certification Providers – Streamline exam development and updates.
• Educators – Build assessments aligned with learning objectives and curricula.
• Corporate Trainers – Develop competency-based training assessments.

Key Features

1. AI-Generated Questions

   • Automatically generate exam questions from structured and unstructured content.
   • Customize outputs by topic, difficulty level, and question format.

2. Source Content Input

   • Upload source material in various formats, including PDFs, Word documents, text files, PowerPoint presentations, Excel spreadsheets, CSV files, and web links.

3. Fine-Tuning

   • Train the AI model using user-specific content to match organizational needs.
   • Leverage historical performance data to optimize outputs.

4. Multilingual Support

   • Provide seamless experiences in English, French, and Spanish.
   • Plan for scalability to additional languages.

5. Export Options

   • Export generated content to multiple formats (PDF, Word, and PowerPoint).

6. Robust API

   • Integrate with external tools and platforms for expanded workflows.

7. SaaS Model

   • Built as a Software-as-a-Service (SaaS) application with multitenancy support for scalability.

AI Workflow

1. Input Source Content – Upload guidelines, topic details, and example materials.
2. Generate Questions – Use AI to create drafts with custom parameters.
3. Review and Refine – Adjust outputs through human review and expert input.
4. Export and Integrate – Finalize questions and export to multiple formats or integrate with external platforms.

Technical Architecture Overview

QuGenAI is a Django-based SaaS application designed to streamline the creation of exam content using AI. It supports fine-tuning, retrieval-augmented generation (RAG) pipelines, and question validation workflows to assist exam creators and educators in generating high-quality, multilingual exam questions.

Core Technologies

• Backend Framework:

  • Django – Handles user management, database interactions, admin interfaces, and multitenancy.

• Database:

  • PostgreSQL – Scalable relational database with JSONB support for flexible data storage.

• AI Integration:

  • OpenAI API or Hugging Face Models – Supports fine-tuning and retrieval-augmented generation (RAG).

• Task Queue:

  • Celery with Redis – Handles background tasks, such as AI processing and report generation.

• File Storage:

  • Amazon S3 – Stores uploaded source content files (PDFs, Word, etc.).
  • Django Storages – Integrates S3 seamlessly with Django.

• Email Services:

  • Mailgun – Handles transactional emails with developer-friendly tools.

• Multilingual Support (Planned):

  • django-parler or modeltranslation – For managing multilingual content.

Architecture Diagram

 User → Django Frontend (Admin + UI) → Django Backend (ORM + Views)
                                   ↓
                      Celery Task Queue (Background Tasks)
                                   ↓
                         AI Models (OpenAI, Hugging Face)
                                   ↓
                             File Storage (S3)

Workflow Overview

1. Source Content Upload

   • Users upload PDFs, Word files, and CSVs via Django’s admin or web UI.

2. AI Processing

   • Content is processed using AI models (fine-tuning and RAG pipelines).
   • Tasks run asynchronously with Celery for scalability.

3. Storage and Review

   • AI-generated questions are stored in PostgreSQL and reviewed via Django’s UI.

4. Export Options

   • Users can export questions to multiple formats (Word, PDF, PPT).

Security

• Authentication:

  • Django Allauth for secure user authentication with email confirmation.

• Authorization:

  • Role-based access control for Admins, Content Creators, and Reviewers.

• Data Encryption:

  • Encrypted file storage and SSL/TLS for secure communication.

Technical Roadmap

Deployment Strategy

Current Setup

• Django is deployed as a Heroku app for simplicity and cost efficiency.
• Celery handles background processing tasks for AI pipelines.
• This setup prioritizes rapid development and proof-of-concept validation.

Future Plans

• Microservices Transition:

  • Introduce an optional FastAPI service for AI workflows when scaling demands increase.
  • Use API gateways or reverse proxies for seamless routing between services.

• Kubernetes Deployment:

  • Deploy to AWS/GCP Kubernetes clusters for scalability and redundancy.

• Internationalization:

  • Add support for multilingual interfaces and localization.

Scaling Goals

• Short-Term Goals:

  • Optimize workflows within a unified app while maintaining low costs.

• Long-Term Goals:

  • Scale AI-specific processing independently.
  • Add API integrations for learning platforms.
  • Introduce multi-region deployments to support international users.

Feature Development

1. Multitenancy Design – Build for scalability while keeping the initial implementation simple for a single developer.
2. AI Fine-Tuning Models – Incorporate techniques like retrieval-augmented generation (RAG) and prompt engineering.
3. Performance Monitoring – Enable feedback loops to refine question quality over time.
4. Integrations – Expand API capabilities for external platform connections.
5. Multilingual Features – Implement language localization with fallback options.

Settings

This project remains largely structured and completely configured per the default Cookiecutter-Django's template settings.

Email Server

In development, it is often nice to see emails sent from your application. For that reason, the local SMTP server Mailpit with a web interface is available as a Docker container.

The mailpit container will start automatically when you run all Docker containers. Please check the cookiecutter-django Docker documentation for more details on how to start all containers.

With Mailpit running, to view messages that are sent by your application, open your browser and go to http://127.0.0.1:8025

Sentry

Sentry is an error logging aggregator service. Sign up for a free account at https://sentry.io/signup/?code=cookiecutter or download and host it yourself. The system is set up with reasonable defaults, including 404 logging and integration with the WSGI application.

You must set the DSN URL in production.

Deployment

The following details how to deploy this application.

Heroku

See detailed cookiecutter-django Heroku documentation.

Docker

See detailed cookiecutter-django Docker documentation.

Contributing

We welcome contributions to QuGenAI!

Before you begin, please take a moment to review our Contributing Guidelines. These guidelines outline our development workflow, coding standards, and expectations for contributions.

Highlights include:

• Branching and Pull Requests – Flexible naming conventions with a focus on keeping the main branch green and deployable.
• Commit Messages – Clear and consistent formatting to ensure an understandable project history.
• Code Reviews – Mandatory reviews to maintain code quality and knowledge sharing.

Your feedback and ideas are also highly encouraged as we continue to evolve QuGenAI’s architecture and features.

Thank you for contributing to QuGenAI!

License

QuGenAI is released under the Functional Source License (FSL) for the first two years. After this period, it transitions to the MIT License.

For more details, see LICENSE or visit fsl.software for an overview of the FSL terms.