I got into programming because I wanted to understand how things work — not just use them.

That curiosity led me to build a compiler from scratch (well, with JFlex and CUP, but I wrote every grammar rule and semantic check myself), train neural networks without touching a single ML library, and evolve AI agents that learn to play games purely through survival.

Along the way I picked up Java, Spring Boot, PostgreSQL, and the habit of not stopping until I understand what's happening under the hood.

Today I build REST APIs and backend systems, and I'm co-founding an AI-powered platform for IT professionals — which means I also deal with architecture decisions, database modeling, and shipping things that actually have to work.

AI & Games — three projects, three different learning paradigms:

• Pong AI — Feedforward neural net + backpropagation, pure Java. The AI learns by adjusting weights from its own mistakes in real time.
• Flappy Bird AI — Genetic algorithm. No error signal — birds evolve through natural selection across generations.
• Dino AI — Neuroevolution with a population of agents competing simultaneously. Faster convergence, more chaos.

Not a coincidence. I built all three to understand why you'd pick one approach over another.

Compilers & Formal Languages:

• Java Compiler — Lexical analysis (JFlex), syntactic and semantic analysis (CUP). Full pipeline from source code to validated AST.

Backend:

• Spring Boot REST API — Layered architecture (Controller → Service → Repository), JPA, OpenAPI docs.
• Appointment System — JDBC, PostgreSQL, OOP design patterns.

Games (without AI):

• Asteroids · Snake

Personal tools (private — built for daily use):

• Apex AI — A personal AI assistant built in Python, inspired by J.A.R.V.I.S. Integrates local LLM inference via Ollama, custom command parsing, and a dark-themed chat interface. My daily driver for automating repetitive tasks and exploring what LLMs can actually do.
• Job Finder — A personal automation tool that scrapes and aggregates job listings based on custom filters. Built because I wanted the jobs to come to me, not the other way around.

Backend   →  Java 17 · Spring Boot 3 · REST APIs · JPA/Hibernate · PostgreSQL
Tools     →  Docker · Git · Maven · Postman · Linux
Also use  →  Python · JavaScript · React

Building something in the AI + education space. Not ready to talk details yet — but the backend is Spring Boot and the problems are interesting.

Open to Java backend opportunities (CLT or PJ).