Projects

P.I.E.N.A.H. (Private Image Evaluation Network on Arduino Hardware) is a Edge AI solution engineered to analyze and monitor spatial occupancy in classroom environments using computer vision techniques.

It leverages a YOLO-based segmentation model to detect and segment specific objects (people, laptops, backpacks) and maps these detections against predefined geometric seat masks to precisely calculate the seats availability. Deployed entirely on an Arduino Uno Q Board, the system ensures high scalability, cost-efficiency, and strict data privacy by executing all visual processing locally, eliminating the need for cloud infrastructure.

Engineered a polyglot persistence platform integrating MongoDB and Neo4j to process and query Big Data from the LDBC Social Network Benchmark. The architecture leverages both document and graph models to optimize storage and retrieval based on the data's semantic structure.

A desktop application for Windows and macOS that uses artificial intelligence to transform any folder on a computer into an efficient and intuitive work environment.