-
PRUNe, University of Florence
- Firenze, Italy
PoC-PoV SpinOff
- Increase the TRL of a SpinOff
- Reduce the Memory and Energy Consumption of AI models
Matteo Pinzauti
I'm a theoretical physicist of complex systems, I create models to simulate various real life scenarios. Currently I focus on reducing waste and energy consumption, hoping to create a better world to live in.
- Email address: matteo.pinzauti98@gmail.com
Work experience
-
Giunti Editore
- Firenze, Italy
R&D for Demand Forecasting in Books
- Forecasting the number of prints for new books with an AI model ad hoc for this case study
- Reduce the waste of copies during the production
- Run a Server with GPU and configure the web API for the Frontend
IT Consultant for Smart Reprint Scheduling
- Automatization of the Reprint Scheduling based on AI prediction
- Waste Reduction
- Supply Chain Optimization
-
Freelance P.IVA: (IT) 07402150481
- Firenze, Italy
- Business or sector: Professional, scientific and technical activities
Scientific Researcher and IT Consultant
- Building ad hoc Research Projects ground up from the Scientific Theory using Data and Scope of the Project
- Forecast of Production via Computer Simulations
- Simulate and Reduce the environmental impact on the resources
- Artificial Intelligence and Machine learning for Big Data
- State of the art Complex Systems models with Stochastic Dynamics, Graph Theory and Clustering
-
HTMX
- Business or sector: Professional, scientific and technical activities
CEO of HTMX
- Develop frontend-backend infrastructure using HTMX
-
Tinetal
- Bastiglia, Italy
- Business or sector: Water supply; sewerage, waste management and remediation activities
IA developer for Energy Optimization
- Energy Efficiency optimization for Hydraulic Systems
- Machine Leaning model for constrained performance optimization
- Data Analysis
-
University of Florence
- Firenze, Italy
Tutor of General Physics
Lectures at Biomedical Engineering
Tutor of Electromagnetism
Lectures at Mechanical and Informatics Engineering
Education and training
Coalition for Sustainable AI
-
Resilient AI Challenge
- France
Main activities:
The Challenge focuses on advancing practical solutions for AI model compression. Participants are invited to develop compressed versions of selected open-source or open-weight AI base models, aiming to achieve the best balance between model accuracy and energy gains.
The models we are working on are:
- Audio-to-text using Voxtral Realtime by Mistral AI
- Image-to-text using Gemma 3n by Google
- Field(s) of study: Artificial Intelligence, Energy Optimization
SIFS
-
SIFS Member
Fosdem
-
FOSDEM
- Brussels, Belgium
Coop, EIIS
-
AI for Food
- Project to create more readable meta-labels to highlight food sustainability and quality
- Field(s) of study: Artificial Intelligence, Retail
- Level in EQF: EQF level 6
- Thesis: Etichetta Trasparente
Linux Foundation
-
Linux Day
- Prato, Italy
Vento
-
Italian Tech Week
- Torino, Italy
SIFS, IUPAP, Triumph
-
StatPhys29
- Firenze, Italy
- Work as a volunteer in the organization of the event and for the coordination of the talks
Università degli Studi di Firenze
-
Master in Theoretical Physics of Complex Systems
- Firenze, Italy
- We created a framework to reduce the Size and Energy Consumption of Pretrained Neural Network models without significant accuracy reduction, making them faster and more Sustainable
- Field(s) of study: Complex Systems
- Final grade: 110 Cum Laude
- Level in EQF: EQF level 8
- Thesis: Spectral Distillation for Neural Networks Pruning
UiT
-
Erasmus + Studio
- Tromsø, Norway
- Solar Energy and Energy Storage
- Pattern Recognition
- Machine Learning
- Field(s) of study: Renewable Energy, Artificial Intelligence
- Final grade: A
- Level in EQF: EQF level 8
- Type of credits: ECTS
- Number of credits: 30
Università degli Studi di Firenze
-
Bachelor in Physics and Astrophysics
- Firenze, Italy
- Generate random fractals with a given Hausdorff dimension using Julia and Python and via machine learning methods we reconstructed the distribution of the parameters from the fractal image
- Field(s) of study: Natural sciences, mathematics and statistics
- Final grade: 108/110
- Level in EQF: EQF level 7
- Thesis: Generation of Fractals and Inverse Problems
MIT Open Courses
-
Computational Thinking
- Field(s) of study: Computer Science
- Level in EQF: EQF level 7
Campus
-
Quantum Mechanics and Applications
- Marina di Massa, Italy
- Field(s) of study: Natural sciences, mathematics and statistics
- Level in EQF: EQF level 6
Language skills
Mother tongue(s)
Italian
Other language(s)
| Listening | Reading | Spoken interaction | Spoken production | Writing | ||
|---|---|---|---|---|---|---|
English |
C1:
Proficient user
|
C1:
Proficient user
|
B2:
Independent user
|
B2:
Independent user
|
C1:
Proficient user
|
|
Spanish |
A1:
Basic User
|
A1:
Basic User
|
A1:
Basic User
|
A1:
Basic User
|
A1:
Basic User
|
Skills
- solar energy
- Microsoft Powerpoint
- machine learning
- analyse big data
- computational physics
- thermodynamics
- perform dimensionality reduction
- statistical modeling techniques
- mathematical physics
- Application experience in MS Excel, MS Word, Python, Stata, The Julia Language and LaTeX
- image recognition
- deep learning
- run simulations
- energy storage systems
- quantum computing
- computer simulation
- CSS & HTML
- HTMX
Volunteering
StatPhys29
- Work as a volunteer in the organization of the event and for the coordination of the talks
Rakettnatt Arctic Electronic Music Festival
Physics Students' Representative
Introduced 6 new modern Master curricula
Scoutmaster AGESCI
Scout Parliament Member
Gruppo Elba APS
- Member of the directive committee
Hobbies and interests
Music
Nature
Snowboarding
Retro Gaming
Driving Licence
AM
B