Author Image

Mattia Sospetti

Mattia Sospetti

Software Engineer at GrottiniLab S.r.L.

Passionate about AI, Computer Vision, and Machine Learning, I have developed AI solutions across various projects with professional experience in multiple European countries. This website showcases my portfolio, showing my professional journey and personal projects I build to strengthen my skills and explore new technologies.

Skills

Experiences

1
GrottiniLab

May 2023 - Present

Macerata, Italy

GrottiniLab uses the latest digital computer vision technologies supported by artificial intelligence (AI), deep learning and machine learning to monitor, analyze and improve the world of Retail and Industry..

Machine Larning Engineer

May 2023 - Present

Responsibilities:
  • Designed and implemented Computer Vision solutions, creating new products or enhancing existing ones with new functionalities.
  • Implemented an embedded tracking solution, proposing innovative firmware and implementing proprietary libraries, reducing costs from €500 per unit to approximately €100.
  • Improved an internal reverse image search solution by adopting updated loss functions and a more sophisticated training task, resulting in a 4% improvement, from 92% to 96%.
  • Ported a tracking solution from C++ to Python to improve long-term reliability and reduce technical debt, and enhanced scalability up to 100 Raspberry Pis simultaneously through automatic monitoring with AWS cloud solutions (RedShift, Lambda, AWS IoT, S3, SQS).

Télécom SudParis

April 2024 - May 2025

Paris, France

Télécom SudParis is a large school that forms general engineers, as well as special engineers and teachers, innovators and entrepreneurs,

Visiting Researcher Intern

April 2024 - May 2025

Responsibilities:
  • Responsible for representing GrottiniLab within the “Trust Rise Project” (Marie-Curie fellowship), participating as a visiting researcher in European universities, where I collaborated with PhDs, and co-authored R&D academic articles in AI and IoT fields.
  • Implemented a real-world simulation system to test MQTT protocols from a cybersecurity perspective.
  • Developed classification neural networks whose accuracy increased by 10% through the adoption of more modern networks and improved training methods.
  • Implemented a backdoor injection on a GAN neural network for risk analysis.
  • paper publication at https://inria.hal.science/hal-04832442/
2

3
Tod's S.p.A.

October 2022 - February 2023

Fermo, Italy

Tod’s is an Italian luxury fashion brand renowned for its high-quality leather goods, including shoes, bags, and accessories.

Data Analyst Intern

October 2022 - February 2023

Responsibilities:
  • Internship done at the Tod’s group. Development of a software architecture, starting from web scrapers to get the data and integrating it into data warehouse and creating Business Intelligence views.

Education

Master's Degree in Computer Science & Automation Engineering
Graduation Mark: 110 Cum Laude out of 110
Taken Courses:
  • OPERATIONS RESEARCH 2
  • NON-LINEAR CONTROL
  • BUSINESS ORGANIZATION
  • COMPUTER VISION AND DEEP LEARNING
  • SOFTWARE CYBERSECURITY
  • ADVANCED PROGRAMMING
  • PROJECT MANAGEMENT FOR ICT
  • AUTOMATION SYSTEMS
  • PROGRAMMING LANGUAGES
  • DATA SCIENCE
  • DRIVES FOR AUTOMATION AND ROBOTICS
  • PREVENTIVE MAINTENANCE FOR ROBOTICS AND INTELLIGENT AUTOMATION
Thesis:
Deep Learning to intercept trends for a fashion company
Supervisor:
Emanuele Frontoni
Bachelor Degree in Computer Science & Automation Engineering
Graduation Mark: 107 out of 110
Taken Courses:
  • MATHEMATICAL ANALYSIS 2
  • BUSINESS ECONOMICS
  • FUNDAMENTALS OF COMPUTER SCIENCE
  • MATHEMATICAL ANALYSIS I
  • GENERAL PHYSICS II
  • GENERAL PHYSICS I
  • LINEAR ALGEBRA AND GEOMETRY
  • ELECTROTECHNICS
  • ELEMENTS OF ELECTRONICS
  • FUNDAMENTALS OF AUTOMATIC CONTROL
  • AUTOMATIC CONTROL SYSTEMS
  • ALGORITHMS AND DATA STRUCTURES
  • PROBABILITY AND MATHEMATICAL STATISTICS
  • ELECTROMAGNETISM FOR INFORMATION TRANSMISSION
  • OBJECT-ORIENTED PROGRAMMING
  • WEB TECHNOLOGIES
  • OPERATIONS RESEARCH
  • TECHNOLOGIES FOR AUTOMATION AND ROBOTICS
  • INDUSTRIAL AUTOMATION
  • OPERATING SYSTEMS
  • MOBILE PROGRAMMING
Thesis:
E-commerce application for the Android Operating System
Supervisor:
Domenico Ursino

Projects

LeetCode top 50 SQL Questions
LeetCode top 50 SQL Questions
Individual April 2025 - June 2025

Answers for LeetCode top 50 SQL Questions (both in SQL and Pandas!)

Data-Analysis
Data-Analysis
Individual February 2025

extensive data analysis (visualization, clustering, classification, tsa) on several dataset examples - Pandas, Statsmodels, prophet, keras, tensorflow, pytorch, scikit-learn, matlplotlib

GAN Backdoor Injection
GAN Backdoor Injection
Main Author November 2024

GAN Backdoor Injection.

SHIELD - assessing security-by-design in federated data spaces using attack graphs
SHIELD - assessing security-by-design in federated data spaces using attack graphs
Contributor April 2024 - June 2024

(Paper publication) SHIELD - assessing security-by-design in federated data spaces using attack graphs

GAN-MNIST-PyTorch
GAN-MNIST-PyTorch
Individual October 2023 - November 2023

Testing training a GAN to generate new digits from MNIST dataset, using PyTorch library.

FashionMNIST-VAE
FashionMNIST-VAE
Individual October 2023 - November 2023

Implementing a Generative AI with VAE, starting from FashionMnist dataset

Univpm_Projects
Univpm_Projects
Individual October 2017 - February 2023

projects made during college years

Accomplishments

Cambridge C1 Level English Certificate

AWS Certified Machine Learning Specialty 2025 - Hands On!

Stanford Verified Certificate for Databases - Advanced Topics in SQL

Stanford Verified Certificate for Databases - Relational Databases and SQL

Reinforcement Learning - beginner to master - AI in Python