Eduvion/EduCore — LLM + RAG-Based Intelligent Educational Automation Platform (MVP - Under Active Development)

What it is

Eduvion/Educore is an intelligent educational automation platform that combines:

LLM-powered content generation (questions, study notes, course content)
RAG (Retrieval-Augmented Generation) so responses come from uploaded institutional/course documents
Role-based portals for Admin, Faculty, and Students
Assessment + evaluation automation with report cards and feedback
Voice-enabled document search + playback (Voice Search + Play)
Many more...

It is designed as a practical “AI teaching assistant + content engine + assessment system” in one workflow.

What we automate (MVP – Under Active Development)

Question Bank Generation (LLM / DB)
Assessment Paper Creation
Online Exam Delivery (Timed)
Auto Grading + Report Cards
Course Design + Study Material Generation
RAG-based Custom Knowledge Assistant (EduCore)
Voice-enabled Search + Audio Playback
More features will be added...

DEMO Snapshot

Full demo is available on request! Email: debashis.das@ieee.org

Authentication + Role-Based Access

Clean Login UI with email/password
OTP Verification during registration/email verification
Three portals:
- Admin Panel
- Faculty Panel
- Student Panel

Admin Panel (platform control)

Admin dashboard provides:

Assessment Generator
Course Design
Resource Library
Lesson Plans
Report Card
Custom GPT (EduCore)
User Management
System Settings
Plus: global “Search from your knowledge base…”

Faculty Panel

Faculty dashboard provides:

1. Assessment Generator
2. Course Design (study materials generation + topic saving)
3. Resource Library
4. Lesson Plans
5. Report Card
6. Custom GPT
7. Knowledge-base search bar for fast retrieval

Student Panel

Student dashboard provides:

1. Assessments & Evaluation
2. Study Materials
3. Lesson Plans
4. Report Card
5. Knowledge-base search bar

Faculty generates a question bank using LLM or custom DB

Faculty opens Assessment Generator

Enters:

Topic (example shown: “Large language model”)
Number of questions
Difficulty level
Question type (Multiple Choice)
Dataset type (example shown: “AI Generated”)

Clicks Generate MCQsSystem shows preview with:

Questions + options
Correct answer indicated
Checkbox selection per question

Faculty can Save Questions into subject/standard categories.

Faculty sets an assessment paper from saved questions

Faculty uses Filter Questions:

Select Standard/Degree (example shown: PhD)
Select Subject (example shown: Large Language Model)

System lists available questions

Faculty checks which questions to include

Clicks Save Assessment Paper

Student takes subject-based exam with timer

Student opens Assessments & Evaluation

Selects Subject to Start Exam

Clicks Start Exam

Exam interface displays:

MCQs in clean format
A visible timer (e.g., “Time Left: 9:56”)

Student submits answers

Auto evaluation + instant result summary

After submission:

Student sees confirmation:
1. Score (example shown: 3/5)
2. Percentage (example shown: 60%)

Report Card generation

Student can open My Report Cards, showing:

Attempt history (Attempt #1, #2, #3…)
Degree + subject
Total questions, correct/wrong
Score + percentage
Status (Pass/Fail)
Short feedback (example: “Practice more” / “Outstanding work”)
Timestamp (“Attempted on …”)

Course Design: Auto study material generation + topic storage

Faculty opens Course Design:
1. Left panel shows course structure (example: Cybersecurity → Introduction → topic list)
2. Main panel generates structured content for a topic (example: “Understanding Security and Privacy”)
3. Faculty can check/select topics and Save Selected Topics

(Custom GPT): RAG training with uploaded files

Faculty opens Custom GPT (EduVion) and sees:
A) Upload RAG files
1. Select Topic (example: Cybersecurity)
2. Choose files
3. Click Upload
4. System shows upload status:
  1. Uploaded / Skipped / Failed
B) View uploaded RAG file summary
A table shows:
1. Subject
2. Original filename + stored filename
3. File type (PDF etc.)
4. Chunks count
5. Upload count
6. Images extracted
7. First/last uploaded timestamps
8. First/last image timestamps

Document-grounded search + Voice + Playback

On the same EduVion page:
1. User enters query in Search Documents
2. Clicks:
  1. Search (text query)
  2. Voice Search (speech input)
  3. Play (audio playback of response)
3. Results appear under “Results Given by EduVion”

Why built it (problem + motivation)

Modern institutions struggle with:

Too much manual effort to create question banks, assessments, lesson plans, and study material
Difficulty keeping learning content aligned with syllabus + instructor standards
Students needing instant, accurate learning support from trusted course content (not random web answers)
Lack of integrated tools: content creation, exam delivery, evaluation, and analytics are often disconnected

Eduvion solves this by making course delivery automated, consistent, and scalable to keep control in the hands of faculty and admin.

Impact

60–80% reduction in academic content creation Time
Faster assessment cycles
Zero manual grading
Scalable AI teaching assistant without hiring costs
Trusted and hallucination-resistant AI responses
Standardized assessments across departments
Curriculum-aligned content generation
Data-driven student performance insights
Improved accreditation readiness
Designed for future expansion into agentic AI and adaptive learning systems
Long-term: Eduvion enables institutions to shift from manual, instructor-centric workflows to intelligent, scalable, learner-centric ecosystems for the next generation of AI-driven learning and assessment.

Page updated

Google Sites

Report abuse