Lorenzo Epifani
Machine Learning Specialist & Developer
📍 Pavia, Italy
📧 lor.epi.job@gmail.com
"Hi! My name is Lorenzo and I was born in Brescia in 1994. Computers have been part of my life since I have memories, feeding my desire to learn, work and collaborate with other people. The command line is my comfort zone."
Work Experience
Machine Learning Researcher & Developer
Politecnico di Milano / A.A.G. Stucchi (PhD Intern)
05/2024 - Ongoing
At AAG Stucchi, I am carrying out my PhD project, which focuses on designing and implementing time series forecasting models. In addition, I oversaw the company's migration from on-premise infrastructure to Azure, and I designed and implemented a custom ETL system that continuously feeds their data to a trusted external third-party service.
Python
Pytorch
Qdrant
Azure
Docker
Databricks
Darts
Neuralforecast
Machine Learning Researcher & Developer
Università del Salento
04/2022 - 12/2023
Focused on deep learning and classic methods for computer vision, implementing them in our core project regarding remote sensing (TeBaKA national project). I implemented algorithms for image processing/extraction from Sentinel2. Produced 3 publications during this period, despite having no prior academic experience.
Python
Sentinel2
UAV imagery
Pytorch
Typescript
Node.js
Bash
Rasterio
Docker
Machine Learning Researcher & Developer
Sorint.tek
06/2021 - 04/2022
Developed a segmentation and anomaly detection system for defective components produced in an industrial plant of a well-known world-leader italian automotive components company. Used both classic and deep learning methods. Hired as a consultant with supervision from the Computer Vision Laboratory of the University of Pavia.
Pytorch
Skimage
Sklearn
Opencv
Docker
Bash
Git
Backend Developer
IQUII ex Juniper-zs
06/2020 - 06/2021
Involved in different developing activities: APIs for backends, cloud-middlewares for API adaption from desktop to mobile front end, minor frontend interventions, database design/maintenance. Hands-on experience with managing AWS cloud services.
MongoDB
Redis
Jira
Javascript
Typescript
Node.js
Angular
Flask
PostgreSQL
AWS (Lambda, KMS, API Gateway, S3, EB)
Cloud Developer (Intern)
BeSharp
09/2019 - 03/2020
M.Sc intern. This company is a leader in native cloud development and migration. During my thesis activity, I studied and investigated possible enhancements of a BeSharp product operating in several cloud computing environments (AWS, GCP and Azure).
Python
SAML
OAuth
Azure
GCP
AWS (Lambda, API Gateway, KMS, SQS, EC2, S3, DynamoDB, Rekognition, CloudFormation)
Skills & Competencies
Languages
- Bash, Python
- Javascript, Typescript
- Latex, HTML, CSS
- GDScript
- C++, Java (High School/Academic)
Frameworks & Libraries
- Pytorch, Rasterio, Opencv
- Darts, Flask, Sklearn, Skimage
- SentinelSat, Pandas, Numpy, Plotly
- Beautiful Soup, Node.js, Express.js
- Vue.js, Axios
Databases & Cloud
- MongoDB, PostgreSQL, MySQL
- Redis, GraphQL, Qdrant
- Docker, Azure, AWS
- OAuth, SAML, Jira
Soft Skills
- Problem solving
- Interpersonal skills
- Emotional intelligence
- Proactivity & Fast learner
- Teamwork & Independence
Topics of Interest
- Time series forecasting
- Computer vision & Satellite imagery
- Statistics & Language models
- Cloud computing
- Software engineering
Publications
[1] TEBAKA: Territorial Basic Knowledge Acquisition. An Agritech Project for Italy: Results on Self-Supervised Semantic Segmentation
IEEE ISCC 2023
[2] Autonomous Intelligent Remote Sensing System for Military Threat Object Detection
Workshop: IEEE TechDefence 2023
[3] Deep learning solutions for agroforestry UAV imagery
Journal: Smart Agricultural Technology
[4] Toward Operational Scenario Awareness: A UAV Swarm-Based Framework
Workshop: IEEE TechDefence 2025
[5] Long-term forecasting: The Good, the Bad, and the Ugly
Journal: Computational Intelligence Magazine (Accepted)
[6] From Zeros to Insight: Univariate Forecasting with Distributional Fidelity in Zero-Inflated Time Series
Conference: TBD (in progress)
[7] VETT: VectorDB-Enabled Transfer-Learning for Time-Series Forecasting
Conference: AI-MLSystems
[8] Harnessing Large Language Models for Data Mesh Adoption in Multi-Organizational Data Ecosystems
Conference: IFKAD 2025
[9] Generative AI in Data Management: Enhancing Data Quality Control for Business Data
Conference: IFKAD 2025
Education
PhD Researcher
Ongoing
Politecnico di Milano
I am exploring the world of foundation models and time series forecasting and how to integrate them.
Stack: LaTex, Python, AWS, Azure, Darts, Neuralforecast, Ray, Databricks.
Stack: LaTex, Python, AWS, Azure, Darts, Neuralforecast, Ray, Databricks.
II Level Master
01/2021 - 01/2022
Università del Salento
Explored areas of data science, from math and classic/statistical learning theory to modern deep models. The final project was a real case study commissioned during studies.
Thesis Stack: Pytorch, Opencv, Sklearn, Skimage, Docker.
Thesis Stack: Pytorch, Opencv, Sklearn, Skimage, Docker.
Master of Science (Computer Engineering)
2017 - 2019
Università di Pavia
Centered on networking, software engineering, operational systems, computer vision, AI principles. Thesis in software engineering with an internship in a cloud expert company.
Bachelor
2014 - 2017
Università del Salento
First approach to data science during thesis in Operative Systems. Diverse course of study including electronics and electromagnetic fields.
Thesis Stack: Hadoop, Apache Nutch, Apache Solr, Java, LaTex.
Thesis Stack: Hadoop, Apache Nutch, Apache Solr, Java, LaTex.
Certifications & Licenses
- Professional engineer license: Score 8/10
- IRDTA 2023: 9th International School on Deep Learning
- IRDTA 2024: 10th International School on Deep Learning
Spoken Languages
| Language | Level |
|---|---|
| Italian | Native |
| English | C1/C2 (Reading C2, Writing C1) |