/'aleksa ra'dojit͡ʃɪt͡ɕ/

Hello there 👋 (I hope you got that reference).

I am a software engineer with a background in data science and machine learning, but my focus has shifted toward building production systems. I am particularly interested in the intersection of backend engineering, AI/LLM applications and infrastructure, with a strong interest in software architecture and system design.

• Building a reproducible VM provisioning pipeline with Packer, Terraform, libvirt and Ansible
• Operating a self-hosted homelab with 20+ services and centralized observability across multiple devices
• Maintaining a reproducible Debian workstation setup through automated installation, system bootstrapping and version-controlled dotfiles
• Building Python backend systems and production-oriented AI/LLM applications

Backend: Python, FastAPI, PostgreSQL, SQLAlchemy

AI / Data: PydanticAI, LangGraph, PyTorch, scikit-learn

Infrastructure: Linux, Docker, Ansible, Terraform, Packer, CI/CD

Observability: Prometheus, Loki, Grafana

A self-hosted Docker environment running 20+ services. Infrastructure is provisioned with Packer, Terraform and libvirt, while Ansible manages system configuration. Renovate automates dependency update proposals, while CI/CD automates validation and deployment to the production environment. Centralized observability across multiple devices using Alloy, Prometheus, Loki and Grafana, with private networking through Tailscale.

Automated Debian Linux installation and system bootstrapping using preseeded installation media, shell scripts and Ansible. The project provisions a reproducible system from a clean installation while securely handling secrets throughout the process.

An end-to-end machine learning project covering data collection from scraped second-hand car listings, persistent storage, feature engineering and model development with hyperparameter optimization. The project demonstrates the full workflow from raw external data to trained predictive models.

A Java client-server desktop application based on TCP sockets. The project covers the design of a distributed application, client-server communication and server-side coordination of shared application data, with detailed documentation and UML diagrams covering system requirements, design and development.

A full-stack aquaristic e-commerce application developed collaboratively using React and Laravel.