Curriculum Vitae

About me – Summary

Professional software engineer and computer science student with an eye for detail, high level of diligence, and a knack for tinkering. I’m deeply invested in mathematics, which is the root of all areas I take interest in. I’m a keen and thorough researcher who loves materializing abstract ideas, mostly through software projects. Some of my current interests are image processing and computer vision, parallelization and concurrency (HPC, SIMD), computer architecture and general systems programming and low-level software engineering (C/C++), statistical learning (ML/DL), and more recently, Bayesian modeling.

Education

• ECTS credits: 60
• Most impactful courses: Deep Learning, Machine Learning, Big Data Systems, Mathematical Methods for AI/ML.

• ECTS credits: 240
• Grade average: 9.8/10
• Most impactful courses: Algorithm Design and Analysis, Computer Vision, Internet of Things, Distributed Systems, Operating Systems, Parallel Systems (HPC & SIMD), Object Oriented Design, Algorithms and Data Structures, Computer Networks, Computer Organization & Architecture, Database Systems.
• Extracurricular activities: Two-time medal winner (2ⁿᵈ place 2023 Ohrid, 3ʳᵈ place 2024 Trebinje) at international student competition in C++ and Object-Oriented Programming (Elektrijada).

Selected Skills

Programming Languages: C, C++ (up to C++20), Python, R, Bash, C#/Java

Scientific/Statistical Computing: NumPy, SciPy, OpenCV, PyTorch, PyMC, scikit-learn, Eigen

High-Performance Computing & Profiling: OpenMP, SIMD (AVX/NEON), Multi-threading (Intel TBB), CPU Profiling (perf, Valgrind, Google Benchmark), Cache Optimization/DOD (Cachegrind)

Tools & Infrastructure: CMake, GCC/MSVC, Linux (Arch/Debian), Docker, Git, GTest, Quarto

Working Experience

• Actively working on algorithms and architecture for 1D barcode decoder libraries, utilized in resource-constrained devices – learning about robust and efficient algorithm design for dynamic systems.
• Performed thorough unit and integration testing using gtest, plus baseline performance evaluations and hardening of C++ decoder libraries on real image datasets – strengthening data analysis/reporting and memory/time performance profiling.
• Designed and implemented robust pattern matching algorithms “from scratch”, in mature C/C++ codebases of Zebra’s 1D barcode decoding libraries – learned about fuzzy pattern matching applied to barcode decoding in noisy environments.
• Additionally: Learned how to use tooling for C/C++ projects, e.g. CMake, Doxygen. Got exposed to AIDC/barcode standards (GS1 General Specifications, ISO/IEC 15420).

• Designed and worked on development of a service-oriented, modular server-side project, used as a back-end for a SIEM Web application, to automate cybersecurity assessment, using Node.js ecosystem – learned security concepts, and worked on implementing service oriented back-end architecture.
• Designed and developed custom backend framework based on Express – learned about “low-level” details of REST APIs, and network programming.
• Developed data visualization dashboards for cybersecurity assessment evaluation, and risk/threat reporting – learned about good visualization practices, automated reporting, and security audits.
• Performed data analysis for cybersecurity vulnerability assessment reports – distilling a large corpus of system-generated records into high-level, presentable reports.
• Additionally: Learned to work in a distributed, remote, and flexible work environment, strengthened my soft skills. Became familiar with security frameworks (NIST CSF 2.0, NIST SP 800-53, CIS Controls v8, ISO/IEC 27001).

• Worked on an app used for making and testing nuclear segmentation pipelines - learned basics of image processing and computer vision, with emphasis on developing/tweaking implementations of well-known algorithms (using SciPy, scikit-image, OpenCV, …); did GUI programming in Python.
• Made a package for automated generation of TOML configuration files through Python code, with an approach heavily relying on dataclasses and tomlkit library – first real-world Python project; learned about TOML format, solidified understanding and knowledge of using Poetry ecosystem.
• Additionally: Learned to completely rely on using terminal and CLI tools, which changed my workflow and effectiveness; solidified Linux knowledge, tried out Manjaro as my daily driver and got used to Arch-based tooling (pacman).

Selected Projects

• Description: Real-time 2D state estimation and tracking system for a drone experiencing physical process noise (wind gusts) and measurement noise (noisy GPS sensor).
• Outcome: Implemented kinematics-based prediction models and simulation-based sequential Bayesian updates (Monte Carlo localization); built a particle filter managing a swarm of N particles to represent position probability distribution; handled real-time resampling to collapse tracking uncertainty.
• Tools Used: C++ (Eigen, OpenMP, GTest, Google Benchmark), Python (NumPy, Matplotlib, Quarto).

• Description: Bayesian inference system designed to restore degraded, blurry, or low-contrast 1D barcode signals back into clear binary patterns.
• Outcome: Formulated the barcode restoration task as a hidden Markov model and binary state estimation problem; modeled physical lens degradation (Gaussian blur) and camera sensor noise as the likelihood function; implemented a neighbor-dependent binary prior to enforce structural barcode constraints; applied Bayesian optimization/inference to reconstruct original sequences.
• Tools Used: C++ (Eigen, GTest), Python (PyMC, OpenCV, NumPy, SciPy, Quarto).

• Description: System for log data aggregation, analysis - pointed towards anomaly detection based on principal component analysis (PCA) – and visualization of results using open source technologies.
• Outcome: Built a log-aggregation and anomaly-detection pipeline for unbounded, out-of-order data sources; applied PCA for anomaly detection and the Drain algorithm for log-template extraction; translated ideas from research papers into a working, enterprise-style pipeline.
• Tools Used: Docker (Compose), Loki, Grafana, ELK stack, Shell scripting, Python (with Poetry).