🎓 Computer Science Student @ York University, Toronto 🤖 Passionate about AI Agent Systems and Java Full-Stack Development, currently preparing for my 2026 Co-op placement. 🌱 Building intelligent applications powered by LLMs, multi-agent orchestration, and RAG pipelines.
💡 Languages: Java, TypeScript, JavaScript, SQL, C 🧰 Frameworks & Libraries: Spring Boot 3, Spring AI, React 19, Next.js 14, Vue.js 3, Tailwind CSS, MyBatis 🗄️ Database: MySQL, PostgreSQL, H2 ⚙️ Tools: Git, Docker, Maven, Postman, IntelliJ IDEA, VS Code 🤖 AI & Agents: Claude API (Anthropic), OpenAI API, Gemini API, SiliconFlow, OpenRouter, RAG, Tool Calling, MCP
| Capability | Description |
|---|---|
| RAG Pipelines | Retrieval-Augmented Generation with vector search for context-aware AI responses |
| Tool Calling | LLM function-calling to invoke APIs, query databases, and trigger backend actions |
| Multi-Agent Orchestration | Coordinating specialized agents via Spring AI for complex task decomposition |
| Prompt Engineering | Structured prompting with system roles, few-shot examples, and chain-of-thought |
| Streaming Responses | Real-time SSE token streaming for responsive chat UIs |
| AI Model Integration | Experience with Claude, GPT-4, Gemini, DeepSeek across multiple providers |
| Project | Description | Tech Stack |
|---|---|---|
| ✨ Aura Quiet Living | A production-ready e-commerce platform powered by Spring AI Agents. Features an intelligent customer service agent with RAG, tool calling for order management, role-based access, and a modern React frontend. | Spring AI, Spring Boot, React 19, TypeScript, MySQL, OpenAI |
| 🧠 Knowledge Base (Lightweight RAG) | A lightweight RAG knowledge base demo with FastAPI + Streamlit + ChromaDB + FastEmbed. Features semantic search, streaming Q&A with graceful fallback, and a framework-agnostic Agent Skill module. No GPU required. | FastAPI, Streamlit, ChromaDB, FastEmbed, DeepSeek, Python |
| 💪 FitLogic | An AI-driven fitness ecosystem with a personalized workout plan generator and an interactive AI coach chat. Uses DeepSeek-V3 via SiliconFlow with Spring AI. Features Google OAuth, email OTP, trainer marketplace, and gym buddy matching. | Spring Boot, Spring AI, Next.js 14, PostgreSQL, DeepSeek, Azure |
| 🎓 York Academic Portal (YAP) | A full-stack university management system featuring AI-powered assistance (Gemini), role-based access control (RBAC), and data visualization. | Spring Boot, Vue.js 3, MySQL, Gemini API |
| 🪄 Gemini Magic Wand | A productivity Chrome extension that upgrades raw inputs into expert structured prompts via Gemini API. Features 4 specialized modes and a glassmorphism UI. | JavaScript, Chrome Manifest V3, Gemini API |
| 🎮 Roco Kingdom CLI | A console-based RPG game built to practice Object-Oriented Programming and modular Java design. | Java, OOP |
Claude Certified Architect — Anthropic Foundations · Valid May 2026 – Nov 2026 · Verify
- Agentic AI: MCP (Model Context Protocol) and multi-agent workflows with Spring AI
- LLM Internals: Context window management, embeddings, and vector similarity search
- Cloud Engineering: Docker Containerization & CI/CD Pipelines on AWS EC2
- System Design: Scalable Architecture & Database Optimization
📧 Email: [email protected] 🔗 LinkedIn: linkedin.com/in/johnny-wang-652820337 📂 Portfolio: github.com/johnnywang-byte
⭐️ "Build agents that think, systems that scale."