CV
Academic CV and selected experience.
Contact Information
| Name | Giorgio Bono |
| Professional Title | AI & Robotics Engineer |
| bono.giorgio.02@gmail.com | |
| Phone | +393319784700 |
Professional Summary
Data Science and AI for Computer Vision and Robotics dual-degree Master’s student with strong analytical and problem-solving skills. Motivated to contribute autonomously and collaborate effectively in team environments.
Experience
-
2026 - Present Stockholm, Sweden
Master's Thesis Student
KTH Royal Institute of Technology
Department of Robotics, Perception and Learning (RPL)
- Working on multi-task reinforcement learning and task representations for robotics.
- Currently working on skill discovery via Vision-Language Models (VLMs).
-
2024 - 2025 Turin, Italy
Perception and Computer Vision Engineer
Squadra Corse DRIVERLESS | PoliTO
- Worked on LiDAR-based odometry for real-time localization in an autonomous vehicle.
- Focused on point cloud alignment and system integration with ROS2, Python, and C++.
-
2025 - 2025 Turin, Italy
AI Workshop Professor
Synesthesia
- Designed and delivered workshops introducing artificial intelligence concepts to elementary school educators.
- Taught practical uses of accessible AI tools, including image generators and chatbots, for classroom integration.
-
2024 - 2025 Turin, Italy
Computer Science Workshop Educator
Synesthesia
- Conducted interactive workshops for children aged 7 to 11.
- Taught fundamental programming concepts using Scratch, Sphero indi, and micro:bit with the Maqueen robot.
-
2025 - 2025 Turin, Italy
UROP Researcher
Politecnico di Torino
From Natural Language Text to Source Code
- Researched Text-to-SQL approaches with small and medium-sized language models up to 15B parameters.
- Worked on the BIRD benchmark and experimented with AdalFlow for optimization and evaluation.
Education
-
2026 - Present Stockholm, Sweden
Master's thesis
KTH Royal Institute of Technology
Multi-task Reinforcement Learning and task representations for robotics
- Conducting Master’s thesis within the Department of Robotics, Perception and Learning (RPL).
- Research focus on multi-task reinforcement learning and task representations for robotics.
- Currently working on skill discovery via Vision-Language Models (VLMs).
-
2025 - Present France
Master M2
Université de Lorraine
Intelligence Artificielle et ses Applications en Vision et Robotique (IA2VR)
- Double-degree Master’s program in Artificial Intelligence, Computer Vision, and Robotics.
-
2025 - Present France
Cursus Ingénieur 3ème année
CentraleSupélec, Université Paris-Saclay
Exchange program
- Relevant coursework: Reinforcement Learning, GPU Programming, Advanced C++, Deep Natural Language Processing, Machine Learning, Deep Learning, Statistical Models, Statistical Learning, Software Engineering.
-
2024 - Present Turin, Italy
M.Sc.
Politecnico di Torino
Artificial Intelligence
- Relevant coursework: Machine Learning and Deep Learning, Numerical and Stochastic Optimization, Distributed Architectures for Big Data.
-
2021 - 2024 Turin, Italy
B.Sc.
Politecnico di Torino
Computer Engineering
- Relevant coursework: Algorithms and Data Structures, Object-Oriented Programming, Applied Electronics.
Projects
-
KTH Master's Thesis - Multi-task Reinforcement Learning
Ongoing thesis within KTH RPL on multi-task reinforcement learning and task representations for robotics, currently focused on skill discovery via Vision-Language Models (VLMs).
-
ZooCAM Challenge
First-place solution for the CentraleSupélec Deep Learning Kaggle challenge on plankton image classification with 1.2M samples and 86 classes.
-
Squadra Corse LiDAR Odometry
LiDAR-based odometry project for real-time autonomous vehicle localization, focused on point cloud alignment and ROS2 system integration.
-
UROP Text-to-SQL with Small LLMs
Research project on Text-to-SQL methods with small and medium-sized language models on the BIRD benchmark.
-
Reinforcement Learning in a Multi-SOM Architecture
Research project developed at CentraleSupélec to control a non-Markovian dynamic system, a vertical rocket, in a continuous environment using connected CXSOM maps.
-
6D Pose Estimation - Driverless Application
Developed a deep learning model for 6D object pose estimation based on GNN and PointNet++, using YOLO and ResNet on the LINEMOD dataset and testing on a simulator for real-time perception.