☕ — Welcome to the workshop. Pull up a chair, the code is still brewing. — ☕
public class TuanTu extends Developer {
public String location = "Hồ Chí Minh City, Việt Nam 🇻🇳";
public String role = "Backend Developer — Java & Spring Boot";
public String stack = "Spring Boot · Laravel · React (Vite)";
public String brewing = "Production-grade APIs & scalable systems";
public String exploring = "System Design · Python · Distributed Systems";
public String openTo = "Remote-first opportunities";
public String[] languages = { "Vietnamese", "English", "French" };
@Override
public String getPhilosophy() {
return "Write backend that holds. Build systems that scale. Ship code that lasts.";
}
}I lean backend — architecture, APIs, data modelling — but I also craft the React interfaces that sit on top of them. Full stack in the truest sense: I own both sides of the counter.
|
Full-Stack Cinema Booking System A production-grade ticketing platform built around correctness and concurrency. Race-condition handling baked in from day one.
|
Production-Grade Blog Platform Built end-to-end: DB schema, UI/UX design, API, and deployment. Documented every architectural decision from zero.
|
|
B2B Order & Fulfillment Platform A backend-heavy platform tackling the complexity of business ordering flows, inventory management, and multi-role access control.
|
Todo App — Engineered Beyond Basics A humble todo app used as a testbed to push Laravel + React patterns — query optimisation, algorithmic thinking, scalable architecture.
|
|
Structured Java Learning Path A disciplined, documented journey from Java fundamentals to advanced patterns — built for depth, not speed.
|
|
2026 is not about collecting new stacks. It's about finishing what's started — and building three projects that actually push the boundary.
Good coffee isn't rushed. Neither is good code.
Before shipping anything new, the existing projects deserve proper care:
- cinebook: Complete Blade → React (Vite) migration, write README with architecture notes
- journo-blog-platform: Deploy to production (VPS / Laravel Forge), document deployment flow
- b2b-order-fulfillment-system: Clean up service layer, add API documentation (Swagger or Postman)
- todolistic-app: Document query optimisation decisions, add README
- java-algorithms-practice: Reach 75+ exercises, cover concurrency & design patterns
⭐⭐ fintrack-expense-api — Consolidate Spring Boot
Known ingredients. Master the recipe.
The difficulty here isn't the code — it's the discipline of doing Spring Boot properly for the first time end-to-end.
- Spring Boot REST API: expense CRUD with clean service/repository layers
- JWT auth with Spring Security — understand the filter chain, not just copy-paste
- Multi-currency support via external exchange rate API
- Scheduled email reports with Spring Scheduler
- Swagger / OpenAPI documentation — fully testable from the browser
- Connect to a React (Vite) frontend — first complete Java full-stack project
Spring Boot Spring Security JWT MySQL React
⭐⭐⭐ logpilot — Learn Real-Time & Async
New concept on the menu: WebSocket. Takes 1–2 weeks to steep properly.
The challenge is WebSocket + async messaging — concepts not covered in the Laravel/React stack. Needs focused study before building.
- Study WebSocket fundamentals + STOMP protocol (1–2 weeks dedicated)
- Real-time log streaming via WebSocket (STOMP) from Spring Boot to frontend
- Multi-app support with API key isolation per registered service
- Threshold-based alerting: trigger alerts when error rate exceeds defined limits
- Async log ingestion — decouple receiving logs from processing them
- Inspired by ELK Stack / Datadog — document the architectural decisions
Spring Boot WebSocket STOMP Async MySQL
⭐⭐⭐⭐ docuparse-api — The Big Leap
FastAPI + LLM + Docker multi-service. Each piece is new. Together, they're the hardest project on the list.
Three things I haven't shipped together before: Python FastAPI, calling an LLM API, and Docker multi-service orchestration. Each one alone is learnable — combining them is the real challenge.
- Learn Python FastAPI fundamentals (routing, Pydantic models, async endpoints)
- Integrate an LLM API (OpenAI / Gemini) for document parsing & extraction
- Design the document ingestion pipeline: upload → parse → extract → return structured data
- Docker multi-service: FastAPI + database + (optional) queue worker in one
docker-compose - Handle large file uploads gracefully — chunked processing, timeout management
- Write clear API documentation with example payloads for every endpoint
Python FastAPI LLM API Docker PostgreSQL
Three projects. Three difficulty levels. One year to ship them all — properly.
Connect the dots. Go deeper, not just wider.
- Build a full-stack capstone project: React + Spring Boot + MySQL — deployed end-to-end
- Study System Design fundamentals: scalability, load balancing, DB indexing, async queues
- First open source contribution — Laravel, Spring Boot, or a framework I actually use
- Study for AWS Cloud Practitioner certification (understand the infrastructure I deploy to)
- Review all 2026 projects — refactor, document, and write a retrospective on each
- Define 2027 roadmap with clearer specialisation direction
