Studex Platform
Overview
Studex Zone is a comprehensive intelligent educational platform that fundamentally relies on artificial intelligence, offering an experience close to the idea of platforms like NotebookLM and Bytes.
The platform is not dedicated to teachers, classes, or an educational community, and there is no teacher-student system within it. Its primary focus is directed at the individual user, especially the student.
The platform's concept is that the student enters Studex Zone, uploads their educational files, then benefits from intelligent services to analyze and understand the content and convert it into educational tools, ultimately creating complete study experiences using artificial intelligence.
User Type
The primary user is the student, who performs the following tasks:
- Create an account.
- Upload educational PDF files.
- Use artificial intelligence services.
- Save their files and results within their account.
- Review all previous operations they have performed.
- Reuse files and services at any time.
Main Sections of the Platform
The platform contains several main sections, with details as follows:
1) Dashboard
The Dashboard is the main section of the account and serves as the user's complete control center. It includes:
- Account information.
- User data.
- Access to all sections of the platform.
- Activity tracking.
- Usage management within the platform.
2) Services
The Services section is one of the most important parts of the platform and is divided into two types:
First: Helpful Tools
This section includes approximately 70 helper tools, characterized by:
- Working directly within the browser.
- Not requiring login.
- Not requiring heavy server processing.
The goal of this section:
- Provide assistance to users.
- Improve site SEO.
- Attract traffic from search engines.
- Help students with quick tools.
Second: Core Services within Studex
There are approximately 12 major core services. These services are characterized by:
- Working through the server.
- Requiring processing time.
- Relying on artificial intelligence.
- Requiring storage and logs.
Examples of services include:
- Smart Summarization.
- Smart Translation.
- Smart Explanation.
- Smart Analysis.
- Mind Map.
- Create Flashcards.
- Create multiple-choice questions.
- Questions with more than one correct answer.
- Various types of tests and educational questions.
How Artificial Intelligence Works
The platform does not rely on just one AI model. Instead, it uses more than 30 to 40 AI models, operating through:
- Running multiple models.
- Comparing results between models.
- Selecting or combining the best results.
- Outputting the best possible answer for the user.
The goal is to achieve higher quality than relying on a single model alone.
File Processing and Artificial Intelligence
The platform accepts:
- PDF files.
- Images.
In cases requiring text extraction (from images or PDF files), OCR is used via:
- Minister OCR
Its function:
- Extract text from PDF files.
- Extract text from images.
- Prepare texts for AI analysis.
The system then relies on:
- Embedding.
- RAG (Retrieval-Augmented Generation).
The workflow steps are:
Upload the file.
Extract the text.
Split the text.
Create Embeddings for the texts.
Save the Vectors.
When a specific service is requested, search for relevant parts.
Send appropriate information to the AI.
Generate a structured result based on the original file.
3) Storage
The Storage section is user storage, similar to Google Drive. Its function:
- Save educational files.
- Store PDF files.
- Reuse files within services.
The user can:
- Upload a PDF file.
- Save it within the platform.
- Get a direct link to the file.
- Share the link with anyone.
- Allow anyone with the link to view or download the file based on settings.
Important feature: If a user uploads a file to Storage once, they can use it later in any service without re-uploading it from their device.
4) Workspace
Workspace is the section that collects user operations. It is not a place for sharing or collaboration. Its function:
- Save a log of all requests.
- Display operation status.
- Access previous results.
- Re-execute requests.
Workflow:
The user selects a service.
Uploads a file or chooses a file from Storage.
The request is sent to the server.
The user is redirected to Workspace.
The request and its status appear.
Request statuses:
- Processing.
- Completed.
- Ready to view.
After the process finishes:
- The user enters the service page itself.
Service Pages
Each service has its own dedicated page. Example: Tests service: The user enters the service and can:
- View questions.
- Answer questions.
- Finish the test.
- See their score.
The following are saved:
- Time spent on the test.
- Number of questions answered.
- Number of correct answers.
- Final score.
- All previous attempts.
They can retake the test multiple times.
Notebook System
There is a Notebook section within the platform, which is a very advanced Notebook. It resembles a real notebook but with digital capabilities. The user can:
- Write.
- Draw.
- Take notes.
- Record lectures.
- Jot down anything.
- Use it while studying.
The goal: Create a comprehensive digital study notebook within the platform.
Design and User Experience
Significant attention has been given to the mobile experience. The platform works on:
- Mobile phones.
- Tablets.
- Computers.
The design is fully responsive and adapts based on:
- Screen size.
- Screen dimensions.
The platform is designed to be excellent on mobile, and includes:
- Dark Mode.
- Light Mode.
The goal is to make the platform comfortable for the eyes while studying.
Web App
Studex is a Web App. The site can be converted into an application on the device.
Frontend-Backend Communication
The system relies on the database as a central point. The method is:
- Frontend sends the request securely.
- Backend receives the request.
- Backend executes the process.
- The database is updated.
- Frontend reads data from the database.
The server is closed, and no server-specific information is contained within the Frontend, nor does the Frontend contain server logic.
Database and Backend Structure
The entire project relies on Supabase and uses:
Supabase Authentication
For account management, such as:
- Google Sign-in
Supabase Storage
For storing user files.
Supabase Database
For storing:
- Users.
- Account data.
- Services.
- Results.
- Activities.
Supabase Vector
For storing:
- Vectors resulting from Embedding.
- RAG data.
Supabase is run:
- Self Hosted
on a dedicated VPS for the project.
Payment
Uses:
- Dodo Payment
Used for:
- Subscriptions.
- Paid plans.
- Payment processing.
Notification System
There is a notification system within the platform, not Push Notifications for mobile. There is a notification button within the interface.
The user is shown:
- Service expiration.
- Request completion.
- Success or failure of an operation.
- Payment status.
- Any updates related to operations.
User Data Retention
After login, everything is saved to the account, including:
- Files.
- Services used.
- AI results.
- Tests.
- Answers.
- Scores.
- Logs.
Project Final Goal
The goal is to create a comprehensive intelligent educational platform that helps the student to:
- Upload educational files.
- Analyze content.
- Understand lessons.
- Create summaries.
- Create questions.
- Create tests.
- Create mind maps.
- Save files.
- Track progress.
- Use AI as a complete educational assistant.
Studex Zone is an AI-powered educational environment that combines storage, analysis, content creation, testing, and studying all in one platform.
Problem
Students face difficulty in converting educational PDFs and images into organized study content, as the process requires using multiple tools and steps such as text extraction, analysis, question generation, and saving results, making the experience inconsistent and difficult to manage.
Solution
Studex Zone provides a single platform that allows students to upload their files and process them with AI to convert them into summaries, explanations, questions, tests, and review materials, while saving all results within the user's account for later reference.
Highlights
AI-Powered Educational Platform
Studex Zone is an individual educational environment for students, fundamentally based on AI to convert learning files into educational tools and content.
Multiple AI Models
The platform does not rely on a single model, but runs between 30 and 40 models and compares results to select or combine the best answer.
Number of Core Services
The platform includes approximately 12 major core services that run via the server and require processing time, storage, and logs.
PDF and Image Processing Pipeline
The platform accepts PDF files and images, and uses OCR to extract text when needed to prepare content for AI analysis.
- RAG (Retrieval-Augmented Generation)
Embedding and RAG Architecture
The system relies on Embedding and RAG to split text, create Vectors, and search for relevant parts before generating the result.
Cloud File Storage
The Storage section is similar to Google Drive for saving educational files and reusing them, with the free account providing 50 MB of storage space.
Request Results Log
Workspace keeps a log of all requests and their statuses (Processing, Completed, Ready to view) with the ability to refer back to previous results.
Dedicated Service Pages for Tests
Each service has its own page. For example, the tests page displays questions, allows solving, calculates scores, and saves time, number of correct answers, scores, and attempts.
Notebook for Writing and Drawing
There is a Notebook within the platform that resembles a physical notebook with digital capabilities including writing, drawing, note-taking, and recording lectures within the study environment.
- Next.js (SSG)
Specific Frontend Technologies
The Frontend relies on Next.js with SSG so that the site's core files are on the server and the browser fetches data from the database.
Backend Technologies and Hosting
The Backend includes Node.js with hosting via Cloudflare Host and a dedicated VPS for the project.
Comprehensive Supabase Database
The project relies entirely on Supabase through multiple components including Authentication, Storage, Database, and Vector to support accounts, files, results, and Vectors for RAG.
Supabase Self Hosted
Supabase Self Hosted is run on a dedicated VPS for the project.
Payment System for Subscriptions and Plans
The system uses Dodo Payment for international payment processing including subscriptions and paid plans.
In-Interface Notifications
There are notifications within the platform (not mobile Push Notifications) that show the user service expiration, request completion, operation success/failure, and payment status.
Features
- Running more than 30 to 40 AI models with result comparison to select or combine the best outputs
- Content analysis structure based on OCR, Embedding, and RAG (Retrieval-Augmented Generation) with Vectors saved for searching text parts during service execution
- Multiple core educational services relying on different Prompts per service to produce results in various ways (such as summarization, explanation, analysis, creating Mind Maps, Flashcards, Quiz Cards, and test types)
- Dedicated Storage for educational files similar to Google Drive with saving a PDF once to reuse later within any service on the platform
- Workspace as an operation layer for requests displaying each process status (Processing, Completed, Ready to view) with the ability to re-execute requests without deleting them while maintaining a complete user usage log
Tech stack
Architecture
Multiple AI Services Ecosystem
The platform offers approximately 12 major core services based on AI, producing educational results such as smart summarization, smart explanation, smart analysis, mind map generation, Flashcards, Quiz Cards, and multiple test types. Each service has its own specific Prompt, making each service's output different and reflecting complexity in content production design. The platform also relies on 30 to 40 models, with model result comparison and selection/combination of the best result to achieve higher quality.
RAG Architecture and File Content Processing
The platform processes PDF and image files. In cases requiring text extraction, Minister OCR is used to extract text from files and images. Then the text is split, Embeddings are created, Vectors are saved, and when a specific service is requested, relevant parts are searched and information is sent to the AI to generate a structured result based on the original file. This sequence demonstrates a multi-stage, interconnected content processing system reflecting clear technical scale and complexity.
Workspace for Request Status and Usage Log
The Workspace section collects user operations by saving a log of all requests, displaying operation status, accessing previous results, and re-executing requests without deleting them. Requests are shown to the user with statuses such as Processing, Completed, and Ready to view, then the user navigates to the service page itself after the process finishes to view the result. This demonstrates a request operation cycle and status management with result reuse at the user interface level.
Advanced Storage Integration for User Files and Vectors
There is a user storage layer within the Storage section similar to Google Drive, saving educational PDF files and allowing direct link creation, file sharing, and managing view/download permissions based on settings. In addition to storage, the system uses Supabase Vector to store Embedding-generated Vectors and RAG data. This integration connects user content and its reuse in services with the vector-based search architecture.
Complete Backend Stack and Dedicated Hosting
The system is built with a Frontend using Next.js with SSG, a Backend using Node.js, hosting via Cloudflare Host, and a dedicated VPS for the project. The platform also relies entirely on Supabase for account authentication, file storage, database, and Supabase Vector, with Supabase Self Hosted running on a dedicated VPS for the project. This reflects a fully integrated architectural design with multiple components and a dedicated hosting environment.
Users
Launched