Giacomo Bellezza Email & Phone Number

Experience in an R&D team, with a focus on Python projects for the design, development, and validation of proprietary scores. Proficiency in applying Language Model (LLM) models for managing and analyzing textual documents, as well as for interacting with them. Demonstrated skills in data collection, analysis, and the management of data flows.

Valbonne, Provenza-Alpi-Costa Azzurra, Francia

Scrum Agile team working on fraud detection and data protection. I develop Spark applications in Scala that work with big data and are periodically processed in the private cloud via Airflow.

Nice Area, France

Valbonne, Provenza-Alpi-Costa Azzurra, Francia

I joined Amadeus as a consultant in the Data Distribution Framework scrum team.Project’s Description: The team is responsible for the near real-time propagation of data from databases to data centers and specific cache levels in the Cloud, ensuring availability and consistency. The team is also in charge of migrating applications and data flows to the Public Cloud.Responsibility: my responsibilities were development and investigation. The objective of the main project was to optimize the flow of data propagation from the central to the cloud. This flow was very sensitive because it was the point of contact between the central databases and those in the cloud. We had to develop a new propagation flow that could handle the data version. In parallel, the team was in charge of both cloud and central caching. These memories have limited capacity but are updated at high speed. I have been involved in the development of the central-cloud data flow for cloud consistency and the development of programmer support tools, optimization, testing. Technologies & Tools: Kafka, Couchbase, Git, Java, python, bash, C, bats, Jira, OpenShift, Docker, Linux,

Perugia, Italia

Context: Master Thesis: Development of AI models applied in Drug Discovery for new immunotherapiesProject’s Description: The aim of the project was to create a Python workflow for the selection of molecules with specific properties in cancer care processes. The obtained model makes use of artificial intelligence technologies combining machine learning models with natural networks.Responsibility: a team of 4 PhD students of Department of Pharmaceutical Sciences, my duty was to design and develop the software. -Design the architecture -Creation of database: collect data from the internet (ChEMBL, DUD-e) clean and join data -Data exploration -Development of the workflow -Training step: select ML and AI models -fitting and optimization -Multimodel stacking implementation -Test step -Analysis of the results.Project presented at the Computational Chemistry Congress,March 28, 2019 to March 29, 2019 - CDDD 6th Meeting, Rome, ItalyTechnologies & Tools: Machine Learning, Deep Learning, Keras(+TensorFLow), ScikitLearn, Pandas, H2O, Knime, Nvidia, Python, csv, Intellij

Perugia Area, Italy

I worked on a data analysis project in the Weedea company. Data were collected from sensors applied to buildings and the goal was to understand the structural health of buildings. My task was to study the anomaly detection techniques and algorithms. In this period I worked with the Django framework, creating REST API related to machine learning.Responsibility: research and software developmentTechnologies & Tools: Anomaly detection, SQL, Short Time Fourier Transform, Django, HTML, CSS, Python

Laurea Magistrale Lm, Computer Engineering, 110/110 E Lode

Activities and Societies: Drug Discovery, Artificial Intelligence Development of AI models applied in Drug Discovery for new immunotherapies

Laurea Triennale, Computer Engineering

Diploma Scientifico, Piano Nazionale Informatica