Salvatore Petrolo

I'm

About

Hi 👋, I'm Salvatore Petrolo, a passionate Machine Learning Engineer with a Master's degree in Artificial Intelligence and Machine Learning, and a Bachelor's degree in Computer Engineering. Originally from the tiny town of Capistrano in Italy, I've always been driven by curiosity and innovation 🚀.

Today, I'm based in Munich, where I specialise in developing cutting-edge AI and machine learning solutions that push the boundaries of technology. With a strong foundation in AI, Computer Engineering and a deep commitment to solving complex problems, I bring creativity and precision to every project I work on.

I love computers 💻, music 🎶 and travelling to discover new places and cultures ✈️.

AI & Machine Learning Engineer.

  • Age:
  • Degree: Master's
  • Email: toredev@outlook.it
  • Freelance: Available

Facts

I've been a computer enthusiast since a was a teen. I learnt to code when I was 16 and I've developed good planning & organizational skills along with an efficient workflow.

Years Coding

Open-source Projects on GitHub

Languages spoken fluently

Years Industrial Experience

Skills

In this section, I showcase a diverse set of technical skills that form the foundation of my work in AI, machine learning, and software development.


Programming

Assembly90%
C100%
C++100%
C#80%
CUDA C/C++100%
Python100%
Swift80%
Java100%
Php80%
Javascript90%
Typescript90%

Machine & Deep Learning

PyTorch

TensorFlow

Transformers

LLamaIndex

LangChain

CoreML

MLC LLM

ONNX

JAX

LLama Cpp

FastAPI

Flask

Weights and Biases

PIL

OpenCV

Cloud Services

Microsoft Azure

Google Cloud Platform

Amazon Web Services


Big Data Management

MongoDB

Apache Spark

Apache Kafka

MapReduce

Apache Storm

Operating Systems & Shells

Linux + Bash/Zsh

Mac Os + Bash/Zsh

Windows + PowerShell

Resume

Here is my professional resume, where I outline my journey through the dynamic fields of AI, machine learning, and software engineering.

Summary

Salvatore Petrolo

AI & Machine Learning Engineer

Machine Learning Engineer with a strong software engineering foundation, specializing in developing, fine-tuning, and deploying Large Language Models (LLMs) at scale for edge and cloud environments. Skilled in pre-training, supervised fine-tuning (SFT), and Reinforcement Learning from Human Feedback (RLHF) to optimize model performance and relevance. Expertise in transformer-based architectures, Retrieval-Augmented Generation (RAG), and Computer Vision.

  • Munich, Bavaria, DE
  • +39 388 7808493
  • toredev@outlook.it

Education

Master's Degree: Artificial Intelligence and Machine Learning

2020 - 2022

University of Calabria, Arcavacata, IT

Thesis: "Deep Anomaly Detection in ECG Signals to Detect Arrhythmias".

Degree Score: 110/110 cum Laude and academic mention.

Bachelor's Degree: Computer Engineering

2016 - 2020

University of Calabria, Arcavacata, IT

Thesis: "Object Oriented Data Language: a language for devoloping dynamic data collection web app".

Degree Score: 110/110 cum Laude.

High School: Techinical Institute for Computer Science

2011 - 2016

IIS Vibo, Vibo Valentia, IT

Score: 100/100.

Professional Experience

Machine Learning Engineer

2023 - Present

Knowlix GmbH, Tutzing, Bavaria, DE

  • Improved documents key information extraction accuracy to 99% by fine-tuning a Vision-Language Model (VLM) to eliminate OCR dependency and enhance document understanding.
  • Optimized cost, performance, and security of document intelligence workflows by developing a serverless API with AWS, integrating a model zoo of LLMs on SageMaker, and implementing asynchronous job execution with callback and polling.
  • Achieved 4x speed-up and 98% accuracy in on-device key information extraction by fine-tuning a Small Language Model (SLM) and optimizing it for Apple mobile deployment using CoreML.
  • Increased document summarization relevance and accuracy by 25% by fine-tuning a Large Language Model (LLM) with Direct Preference Optimization (DPO) to align summaries with user preferences and domain-specific requirements.
  • Enhanced multilingual classification accuracy by 10% by conducting transformer encoder pre-training for improved downstream performance.
  • Enabled precise document interaction via natural language queries by building a Retrieval-Augmented Generation (RAG) system and integrating a vector search engine for efficient retrieval.

Acknowledgments

Best Student Award for the 2nd Year Master's program at University of Calabria.

2022

University of Calabria, Arcavacata, IT

Best Student for the 2nd Year Master's program in Computer Engineering at University of Calabria. [link]


Certifications

Cisco CCNA 1: Routing and Switching

2016

Cisco IT Essentials

2015

IELTS: B2

2015

Contact

Location:

Munich, Bavaria DE

Phone:

+39 388 7808493