- Software Engineering @ University of Novi Sad, Serbia
- Passionate about machine learning, computer vision, mobile development
- Strong interest in algorithms, data structures, and practical software engineering
- GPA: 9.59 | Completed Harvard’s CS50 & CS50AI
🏋️ Android Barbell Tracker — Kotlin / TensorFlow Lite
Video-based barbell detection and tracking on Android:
- Models: YOLOv8 nano, SORT tracking
- Features: Barbell selection, motion path visualization, GPU acceleration
- Deployment: On-device inference for mobile apps
🐍 YOLOv8 Barbell Tracker — Python / PyTorch
Fine-tuned YOLOv8 nano model for barbell detection:
- Computer Vision: Object detection and tracking
- Visualization: Motion path analysis
- Export: TensorFlow Lite for mobile deployment
🗄️ NoSQL Engine — C++
High-performance Key-Value storage engine:
- Data Structures: LSM Trees, WAL, Bloom Filters
- Features: Durable storage, configurable compaction, range queries
🤖 GA Black Box Optimization — Python
Applied genetic algorithms to optimize unknown high-dimensional functions:
- Real-valued GA with population-based evolution
- Objective: Minimization of black-box functions
🎮 AI Checkers Bot — Python
Checkers AI using Minimax with alpha-beta pruning:
- Strategic decision making
- Real-time move evaluation
- Traffic Computer Vision: CNN for German traffic sign classification with 95%+ accuracy
- Nim AI: Reinforcement learning agent trained via Q-learning and self-play
- Shopping Prediction Model: KNN classifier for predicting online purchase intent
- BERT Attention Visualizer: Visualizing self-attention scores in NLP models
- Parallel Scraper (C++ / TBB): Multi-threaded web scraping with thread-safe pipelines
- Minotaur Labyrinth Game (C++): Console-based maze game demonstrating OOP principles
Languages: Python · Kotlin · C++ · Java
Machine Learning: PyTorch · TensorFlow · scikit-learn · NLP · Computer Vision
Mobile & Deployment: Android · TensorFlow Lite
Other: Git · Algorithms & Data Structures · Parallel Programming · Software Architecture
