STUDENT PROJECT • RAG PLATFORM
Exam Intelligence Assistant
From Question Papers to Intelligent Answers.
Built by a PBC Student
Panskura Banamali College
BCA & CS Department Specific
An AI-powered academic assistant using Retrieval-Augmented Generation (RAG) to help Panskura Banamali College students in the BCA & Computer Science (CS) Department explore previous-year question papers, find relevant questions, and understand topics with grounded answers.
No hallucinations. Grounded in indexed PBC papers.
RAG Pipeline
END-TO-END
01
Student asks a question
Natural language input
02
Embedding + pgvector search
Semantic cosine similarity in PostgreSQL
03
Past PBC questions retrieved
With year, semester & marks metadata
04
Gemini generates answer
Grounded in retrieved exam context
OUTPUT PREVIEW
Example: What Students Receive
Visual Demonstration
Student Question
"What are the common operations on Data
Structures?"
Concept Explanation
AI-generated, curriculum-grounded
Traversal
Visiting and processing each element exactly
once.
Insertion
Adding a new element at any specified
position.
Deletion
Removing an element and reorganizing the
structure.
Searching & Sorting
Finding elements and arranging data in order.
Merging
Combining two structures into one.
Retrieved PYQs
From indexed PBC exam archive
"What are the various operations that can be performed on different Data Structures?"
"What is stack? Write down the operation of stack in memory."
Data scope: Panskura Banamali College (PBC) • BCA & CS
Department Specific. Exam papers from other departments or institutions are not
indexed.