About Me
I build full-stack applications with Angular and Spring Boot, solve production problems, and experiment with local AI systems. I'm a software engineer at Engineering Software Lab in Serbia. Still in school too - finishing my degree at Metropolitan University, graduating in 2027.
What I actually do
Most of my days are spent building enterprise applications - Angular components, Spring Boot APIs, debugging production issues, code reviews. It's practical engineering work. I've been at Engineering Software Lab since 2022.
The work involves microservices, event-driven systems with RabbitMQ, and various databases (MySQL, MSSQL, Oracle, Redis). What I enjoy most is the problem-solving - finally tracking down a tricky bug or finding an elegant solution to a complex problem.
Working towards AI/ML
Third year of software engineering right now. Taking AI courses, building AI projects on the side. Plan is to finish undergrad in 2027, then do a master's in AI. Not there yet, but working on it.
Side projects are where I explore this.
Project Aeon is what I'm building
now - a local AI assistant with FastAPI, ChromaDB for vector search, and Vue 3 for the interface.
Runs entirely offline using Ollama. There's also an option to use Perplexity's API for internet
results with citations. The goal isn't just building a useful tool, it's understanding how these
systems work by building them from scratch.
What I love about this work
I love fast prototyping. There's something satisfying about taking an idea and turning it into working code in a weekend. Small-to-medium projects are my sweet spot - big enough to be meaningful but small enough to iterate quickly. You learn way more building stuff than watching tutorials.
With local LLMs especially, the possibilities are interesting. What can you build if you have a powerful language model running entirely on your own hardware? How do you design interfaces for AI tools? How do you make them fast, reliable, and private?
The learning process itself is what hooks me. Figuring out how things work under the hood, debugging, reading source code, seeing ideas come to life - this is where I thrive.
How I think about building
Building is the best way to learn. You can read papers and watch talks, but nothing compares to actually implementing something yourself, hitting the edge cases, solving real problems. That's why I focus on projects over theory.
Where I'm headed
The roadmap is clear: finish undergrad in 2027, then pursue a master's in AI. Keep building projects, deepen my understanding of ML fundamentals, and transition from general full-stack work to specialized AI/ML engineering.
Some things I've built
Beyond Project Aeon, I've built a distributed chat system (Go, WebSockets, Redis pub/sub) to understand horizontal scaling, and a password manager to learn about secure encryption. Each project teaches me something new.
Check out all of my projects →
What I'm learning right now
At university, I'm taking AI courses - machine learning algorithms, neural networks, NLP. At work, getting better at Angular and Spring Boot through real-world applications. On the side, diving into local LLM deployment, vector databases, and RAG architectures through Project Aeon.
Let's connect
Always up for conversations about AI/ML, local-first architectures, or interesting projects. Want to discuss technical challenges, share ideas, or just say hi? Reach out:
Currently based in Belgrade, Serbia. Usually found debugging something or prototyping a new idea.