TutorAI — AI Tutorial Assistant & Study Packet Generator

TutorAI — AI Tutorial Assistant & Study Packet Generator

TutorAI

AI Tutorial Assistant & Study Packet Generator • Java 21 • Spring Boot 3.4 • Spring AI • React 19 • Google Gemini

An intelligent educational assistant that transforms uploaded study materials—including PDFs, Word documents, plain text, and photographed notes via client-side OCR—into structured study reviewers, interactive self-assessments, and downloadable formatted PDF & Word DOCX packets with detached answer keys.

Launch TutorAI View on GitHub Key Features

About TutorAI

Studying dense educational material—lengthy textbook chapters, lecture notes, research essays, or historical articles—is often a passive, time-consuming process. TutorAI bridges the gap between passive reading and active recall by instantly converting raw educational content into comprehensive, high-yield study packets in under 15 seconds.

Designed as a 100% stateless web application, TutorAI requires zero user registration and zero database storage. All document processing happens on-the-fly, keeping personal notes private and operational overhead minimal.

Key Capabilities

📄 Multi-Format Document Ingestion
Document Parsing & OCR

Upload PDF files (parsed via Apache PDFBox), Word documents (.docx via Apache POI), plain text files, or snap a photo of physical textbook pages using browser-based client-side optical character recognition (OCR via Tesseract.js).

🧠 Structured AI Review Generation
Gemini 2.5 Flash & Spring AI

Leveraging Gemini's massive 1,000,000+ token context window, TutorAI extracts core concepts, definitions, Multiple Choice questions (with 1 correct answer and 3 plausible distractors), True/False statements, and Fill-in-the-Blank exercises in a single structured pass.

🎯 Interactive Practice Exam
In-Browser Self-Assessment

Test your knowledge immediately in the web browser. The interactive exam provides real-time scoring, instant answer verification, source citations pointing back to the uploaded material, and seamless session caching.

📥 Professional PDF & DOCX Export
Offline Printable Packets

Export high-quality, formatted study packets on demand. Generated PDFs (OpenPDF) and Word documents (Apache POI) feature a clean student practice sheet followed by a clean page break separating the detached Answer Key & Explanations.

🛡️ Privacy-First & 100% Stateless
Zero-Retention Architecture

No database, no user accounts, and no persistent cloud storage. Documents are streamed in memory during processing and discarded immediately, ensuring total privacy for proprietary curriculum and sensitive personal notes.

⚡ Rapid 15-Second Generation
Low-Latency Performance

By eliminating vector database chunking and traditional multi-step RAG pipelines, entire multi-page documents are analyzed holistically with sub-15-second generation times and instant offline download triggers.

Architecture & Engineering Highlights

📦 Single Unified Container

Multi-stage Docker build that compiles the React 19 + TypeScript + Vite + Tailwind CSS frontend and packages static assets directly into the Spring Boot 3 executable JAR, served as a single lightweight container image.

🤖 Spring AI Integration

Built on Spring AI (OpenAI API specification compatibility) connecting directly to Google Gemini's generative AI models with strict schema-conforming structured JSON outputs for reliable question generation.

☁️ Render Free-Tier Optimization

Fine-tuned for constrained container memory budgets (≤ 512MB RAM) using JVM Serial Garbage Collection, explicit heap allocation limits (-Xmx350m -Xms128m), and health check monitoring (/health).

🔄 Automated CI/CD & Deploy Hook

GitHub Actions pipeline running oxlint, TypeScript checks, and Maven unit tests. Upon merging to main, multi-stage containers are pushed to GitHub Container Registry (GHCR) and automatically deployed to Render via deploy hooks.

Technical Features & Tech Stack

  • Java 21 (LTS)
  • Spring Boot 3.4.3
  • Spring AI 1.0.0
  • Google Gemini 2.5 Flash
  • React 19 & TypeScript
  • Vite & Tailwind CSS v4
  • Lucide React Icons
  • Tesseract.js (Client-side OCR)
  • Apache PDFBox 3.0.4
  • Apache POI 5.4.0 (DOCX)
  • OpenPDF 2.0.3 (PDF Export)
  • Docker & GitHub Container Registry (GHCR)
  • Render Web Service
  • GitHub Actions CI/CD

🤖 AI-Assisted Development Transparency

TutorAI was engineered using an Orchestrator–Executor multi-agent workflow pairing high-reasoning primary models with high-speed execution engines to maximize architectural rigor and rapid delivery:

Multi-Agent Workflow Pattern

  • Orchestrator (High Reasoning) — Backend architecture, Spring AI contract modeling, prompt engineering, security boundaries, and code review
  • Executor Subagents (High Speed) — Vite/React UI scaffolding, Tailwind styling, Tesseract.js client OCR hooks, and test fixtures

Development & CI/CD Tooling

  • Google Antigravity & CLI — Agentic workspace planning and task decomposition
  • Opencode — Rapid iterative code generation and validation
  • GitHub Actions & GHCR — Automated build, lint, containerization, and deployment