iOS Engineer | ๐ Blockchain Passionate | Apple Enthusiast ๐ | Gym lover
๐ In love with technology and innovation
I'm part of the Graduate Talent Program at UBS, working as a Software Engineer in the WMPC EMEA Tech division.
My role focuses on developing internal tools to support server and application obsolescence management across the EMEA regionโensuring technological sustainability, lifecycle compliance, and proactive risk mitigation.
PlateHunt is the first app in Italy that tracks and analyzes vehicle registration data to identify the most recent license plates released.
It provides statistical insights and predictive models by leveraging:
Last Update - 01/12/2025: Trying to developing a Python script to check new data every morning (it will run on an Ubuntu Server on AWS)
Swift ยท Core ML ยท Create ML ยท MLX ยท Firebase ยท Firebase Functions ยท Python backend
๐ Features:
- ๐ Real-time search and tracking of recently issued license plates
- ๐ Statistical analysis & prediction of registration trends
- ๐ค Machine learning integration for forecasting using Core ML, Create ML and MLX
- โ๏ธ Backend supported by Firebase + Python server
Currently under testing and optimization before public release.
๐ Fully Decentralized โข Trustless โข Lives Forever on the Blockchain
-
๐ Backend / Server
https://github.com/theblackoreo/HyperShipBattle_Server -
๐จ Frontend Client
https://github.com/theblackoreo/deep-battle-ship
This project implements a fully on-chain battleship game powered by the Solana blockchain, where every in-game action is represented by a transaction. Players interact directly with smart contracts, ensuring transparent mechanics and provable fairness.
- ๐ Fully On-Chain Logic (no centralized server)
- ๐ฐ Decentralized State Management
- โก High-performance and low-cost transactions (Solana)
- ๐ช Token-based Rewards System
- ๐ Each shot is randomly assigned per transaction
- โพ๏ธ Immutable & permanently hosted on-chain
| Layer | Technologies |
|---|---|
| Blockchain | Solana |
| Smart Contracts | Rust + Anchor |
| Frontend | React |
| Architecture | Fully On-Chain, Token-Enabled |
A feature-rich mobile app that revolutionizes accommodation planning for travelers.
Developed using Swift & SwiftUI, Whooliday offers users smart tools to discover and book hotels or apartments using advanced search, real-time tracking, and intelligent insights.
- ๐ Advanced Search + high-quality accommodation previews
- ๐ Detailed price trend charts for strategic planning
- ๐ Real-time price tracking and alerts
- ๐ Suggested destinations with historical climate data
- ๐ Multilingual support (EN, IT, DE, ES, FR, ZH)
| Layer | Tech |
|---|---|
| UI | SwiftUI |
| Structure | MVVM + Combine |
| Services | Firebase ยท Google Maps API ยท Open Meteo API ยท Firestore |
| Design | 3-tier client-server architecture |
๐ Repository: https://github.com/theblackoreo/Whooliday
A decentralized reputation-based routing system for mobile ad-hoc networks (MANETs)
๐ Developed as part of my Masterโs Thesis at Florida Atlantic University (FAU, USA) within the AI & Autonomous Vehicles Lab.
SAHARA (Secure Ad-Hoc Advanced Resilient Architecture) is a decentralized system designed to improve routing decisions, network resilience, and security in mobile ad-hoc networks with minimal overhead. It is specifically engineered for resource-constrained environments, such as autonomous robot and drone swarms.
- ๐ก Routing Protocol Optimization
- ๐ Efficient Information Dissemination
- ๐ Network Integrity & Security
The system autonomously adapts to dynamic conditions, mitigating attacks while preserving performance and energy efficiency.
| Feature | Description |
|---|---|
| ๐ Information Dissemination | SAHARA Sync uses lightweight diffusion to reduce overhead vs. traditional flooding |
| โญ Reputation-Based Evaluation | Nodes are assessed on trustworthiness to improve routing decisions |
| ๐ก Resilience to Attacks | Detects and responds to malicious activity (packet drop, DoS) ensuring performance recovery |
| ๐ค Self-Adapting System | Automatically adjusts to changing conditions to improve efficiency & security |
| ๐งช Simulation | Fully simulated using NS-3 (discrete-event network simulator) |
Ideal for swarm robotics and drone networks, where resource efficiency, decentralized decision-making, and security are critical.
๐ก Key benefits:
- ๐ Improved routing robustness via reputation-based decisions
- ๐ Reduced communication and energy costs
- ๐ Enhanced protection from internal attacks on network integrity
๐ Repository: https://github.com/theblackoreo/smesh
Group project focused on human-based visual quality evaluation of images.
๐ Repository: https://github.com/theblackoreo/image_quality_assessment
๐ Live website: https://siqa.pythonanywhere.com
This project implements a subjective image quality assessment system where users rate visual content to generate quality insights. A full-stack approach was used to create a responsive web platform enabling testing, data collection, and visualization.
- ๐ Subjective rating-based quality evaluation
- ๐ Live web application (PythonAnywhere hosting)
- ๐น Preview video + interactive UI
โ ๏ธ Webhook-based auto-deploy integrated into the repository
Python ยท Jupyter Notebook ยท JavaScript ยท HTML5 ยท CSS3 ยท Flask ยท Webhooks
Designed to evaluate perception-driven image quality using crowdsourced human judgment.
Algorithmic exploration of maze solving using image-based extraction and parallel computation.
๐ Repository: https://github.com/theblackoreo/ACA-Project
This project aims to solve a perfect maze (only one solution) extracted from binary or grayscale images using OpenCV. Two implementations were built:
โ Serial version
โ Parallel version (OpenMPI) with performance benchmark on Google Cloud clusters.
| Feature | Description |
|---|---|
| ๐ผ Image-based maze extraction | using OpenCV (C++) |
| ๐ Perfect maze logic | one entrance, one exit, one unique path |
| โก Parallel algorithm | OpenMPI on multi-core VMs |
| ๐ Performance analysis | speedup & execution timing |
| ๐ฌ Optimized for large-scale data | up to 72M+ pixels |
C++ ยท OpenCV ยท OpenMPI ยท Google Cloud Cluster ยท High-Performance Computing
A high-efficiency pathfinding algorithm designed for large image-based labyrinths with multithreading optimization.