Soojal Kumar
Back to Projects

2026

CampusStudy AI

Live university-focused AI study platform with a Vercel frontend, Render FastAPI backend, RAG-grounded chat, notes, flashcards, quizzes, and study guides.

CampusStudy AI is a live university-focused AI study platform that helps students turn course materials into structured study support. The deployed system uses a Vercel-hosted Next.js frontend and a Render-hosted FastAPI backend, with repo support for an Expo mobile companion, Celery worker pipeline, PostgreSQL with pgvector, Redis, object storage, and provider abstractions for LLMs, embeddings, speech-to-text, and storage.

AIRAGFastAPINext.jsTypeScriptExpoWorkersDocument ProcessingFlashcardsQuizzes
Live Demo GitHub API Docs API Health All Projects

Generated from project structure

Live Full-Stack AI Study Platform

CampusStudy AI study dashboard visual

Vercel Web Frontend

Render FastAPI Backend

API Docs + Health

Celery Pipeline

Highlight

Live Vercel Frontend

Highlight

Render FastAPI Backend

Highlight

RAG-Grounded Chat

Highlight

Notes + Flashcards + Quizzes

Executive Summary

CampusStudy AI is a live university-focused AI study platform that helps students turn course materials into structured study support. The deployed system uses a Vercel-hosted Next.js frontend and a Render-hosted FastAPI backend, with repo support for an Expo mobile companion, Celery worker pipeline, PostgreSQL with pgvector, Redis, object storage, and provider abstractions for LLMs, embeddings, speech-to-text, and storage.

Problem Statement

Students often use disconnected tools for notes, slides, lecture recordings, quizzes, flashcards, study planning, and AI help. CampusStudy AI brings these workflows into one academic platform where uploaded course material can become searchable, structured, and reusable study outputs.

What I Built

Next.js student dashboard

Expo mobile companion

Material upload and processing pipeline

PDF, TXT, Markdown, DOCX, PPTX, audio, and video handling

RAG-grounded chat with citations

Generated notes, summaries, study guides, flashcards, and quizzes

Course, topic, enrollment, and university structure

Admin metrics and processing job visibility

How It Works

A conceptual workflow showing how the project moves from input to processing and output.

Step 1

Student Uploads Material

Step 2

Extract / Transcribe

Step 3

Chunk + Embed

Step 4

Generate Study Assets

Step 5

Ask with Citations

Step 6

Review Notes / Flashcards / Quizzes

Architecture / System Design

A simplified system view of the major project components and how responsibilities connect.

Step 1

Next.js Web Frontend on Vercel

Step 2

FastAPI Backend on Render

Step 3

SQLAlchemy Services

Step 4

Celery Workers

Step 5

PostgreSQL + pgvector / documented Render DATABASE_URL

Step 6

Redis + MinIO/S3 storage support

Step 7

LLM / Embedding / Speech Providers

Technical Implementation

Frontend

  • Next.js App Router
  • React 19
  • Tailwind CSS
  • React Query
  • Recharts dashboard surfaces
  • Vercel deployment

Mobile

  • Expo Router
  • React Native
  • Course and study companion flows
  • Secure local storage

Backend

  • Render-hosted FastAPI API
  • SQLAlchemy 2.0 models
  • Alembic migrations
  • Auth, courses, materials, notes, quizzes, chat, admin

AI / RAG

  • Chunking and embeddings
  • Citation-grounded retrieval
  • LLM provider abstraction
  • Generated notes, flashcards, quizzes, and study guides

Workers / Infra

  • Celery pipeline
  • Redis broker
  • PostgreSQL + pgvector
  • MinIO/S3-compatible storage
  • Docker Compose local stack

Deployment

  • Vercel frontend
  • Render backend
  • FastAPI docs endpoint
  • Health endpoint
  • CORS configured for Vercel origins

Testing / CI

  • pytest backend tests
  • ruff checks
  • web lint/test/build
  • mobile typecheck
  • GitHub Actions CI

Live Deployment

Confirmed deployment links and environment notes from the project repository.

Frontend

Vercel - campusstudy-ai-web.vercel.app

Backend

Render - campusstudy-ai.onrender.com

API Base

https://campusstudy-ai.onrender.com/api/v1

API Docs

FastAPI Swagger docs

Render Startup

Docker startup runs migrations, seeds demo data, then starts Uvicorn

Documented Backend Env

DATABASE_URL, SECRET_KEY, LLM_PROVIDER, GROQ_API_KEY, GROQ_MODEL, CORS origins

Screenshots & Visuals

Real project screenshots and outputs appear first. Where a project has no existing screenshots, the visuals are grounded diagrams or output previews based on the actual project structure.

CampusStudy AI study dashboard visual
DiagramGenerated from project structure

Study Workspace Diagram

Generated visual grounded in the latest repo routes for dashboard, courses, materials, notes, flashcards, quizzes, chat, and admin surfaces.

CampusStudy AI retrieval augmented generation workflow
WorkflowGenerated from project structure

RAG + Citation Flow

Workflow diagram based on the implemented chunking, embedding, scoped retrieval, citation, and chat response services.

CampusStudy AI FastAPI and worker pipeline
ArchitectureGenerated from project structure

Background Processing Pipeline

Architecture visual grounded in the Celery processing statuses for extraction, transcription, chunking, embeddings, notes, flashcards, quizzes, and completion.

CampusStudy AI FastAPI API surface map
ArchitectureGenerated from project structure

API Surface Map

Backend route map based on the current FastAPI API domains for auth, courses, materials, processing, notes, flashcards, quizzes, chat, transcripts, dashboard, and admin.

Live API + Confirmed API Surface

Frontend: https://campusstudy-ai-web.vercel.app
API base: https://campusstudy-ai.onrender.com/api/v1
Docs: https://campusstudy-ai.onrender.com/docs
Health: GET /api/v1/health -> {"status":"ok"}

POST /api/v1/materials/upload
GET  /api/v1/materials/{id}/chunks
GET  /api/v1/materials/{id}/status
POST /api/v1/notes/generate
POST /api/v1/flashcards/generate
POST /api/v1/quizzes/generate
POST /api/v1/chat/threads/{threadId}/messages

Challenges & Solutions

Challenge

Students need notes, flashcards, quizzes, study guides, and AI help from the same course material, but these workflows are usually split across separate tools.

Solution

CampusStudy AI centralizes materials, courses, generated study assets, and chat workflows in one web/mobile study platform.

Challenge

AI study answers need to stay connected to the uploaded lecture or course source material.

Solution

The backend chunks uploaded content, stores retrieval-ready embeddings, and returns citation-grounded chat answers from scoped material, topic, or course context.

Challenge

Long-running extraction, transcription, embedding, and generation work can block the user experience.

Solution

A Celery worker pipeline moves materials through explicit processing statuses while the web app can track jobs and surface generated results when ready.

Results / Impact

Demonstrates full-stack AI product engineering across web, mobile, backend APIs, background jobs, data modeling, and retrieval-based study workflows.

Shows practical education technology thinking by turning raw course material into notes, flashcards, quizzes, study guides, and cited AI assistance.

Uses a scalable architecture foundation with provider abstractions, worker queues, vector retrieval, and production guardrails instead of a single-page prototype.

View Live Demo GitHub Link View API Docs View API Health Back to All Projects