Building applied AI systems — OCR pipelines, real-time computer vision, and backend infrastructure engineered for production.
Computer Science & Engineering student at the University of Nicosia (graduating 2026), working at the intersection of machine learning, robotics, image processing, distributed infrastructure, and production engineering.
Recent work includes: ROS 2 drone swarm systems for disaster relief (A.U.R.A), Cypriot dialect ASR dataset engineering, OCR preprocessing pipelines, real-time face tracking with 3D overlays, backend ETL systems for ML ingestion, and cloud-integrated metadata extraction services.
| Domain | Technologies |
|---|---|
| Languages |     |
| AI / CV |     |
| Robotics |    |
| Backend |     |
ROS 2-based drone swarm system for post-earthquake disaster relief. Deploys 5 UAVs that autonomously establish a WiFi mesh network over collapsed infrastructure, providing connectivity to civilians and first responders. Integrates Gazebo Classic simulation, PX4 SITL autopilot, a log-distance path loss network simulator, and PPO reinforcement learning for intelligent swarm positioning. Achieves 96–99% coverage, 114.7 Mbps aggregate throughput, and 6ms latency across the disaster zone.
ROS 2PX4GazeboPyTorchPPODockerC++Python
Production-oriented OCR preprocessing system using OpenCV, scikit-image, PyTesseract, perspective warping, Hough transforms, and contour detection. Designed for robustness under skew, blur, rotation, shadows, and inconsistent document geometry.
End-to-end ETL pipeline for speech recognition dataset generation — broadcast media extraction, segmentation, normalization, cleaning, metadata generation, and ASR-ready formatting. Built for applied speech model fine-tuning workflows.
Low-latency CV application integrating facial keypoint detection, real-time frame processing, 3D mask alignment, and live rendering pipelines.
Autonomous Robotics & Swarms ██████████████████████░░ 90%
Applied AI Engineering ██████████████████████░░ 88%
Computer Vision Systems █████████████████████░░░ 84%
OCR & Document Understanding ████████████████████░░░░ 80%
Backend Architecture ███████████████████░░░░░ 76%