About Me
Hi, I am Leandro Salemi 
,
I hold a PhD in physics and am currently working as a Data Scientist and ML Engineer
for Helicon Technologies.
I enrolled in university at 16, showcasing a strong aptitude for learning and
problem-solving. At 21, I moved to Sweden to pursue my PhD. Over the years,
I’ve developed a strong technical foundation in machine learning with tools like
Tensorflow, PyTorch, and scikit-learn, alongside expertise in cloud and ML engineering
with AWS, Docker and MLFlow. A dedicated Linux user for over a decade, I bring extensive
experience in Python-based development and version control with Git and GitHub/GitLab.
I thrive in managing and leading research and business projects, leveraging my autonomy,
self-driven nature, and strong planning skills. Fluent in French and English, with
advanced Japanese proficiency (JLPT N2), I love collaborating within multicultural
teams. I’m mobile, adaptable, and excited about embracing new challenges, wherever they
may be.
Feel free to contact me if you want to know more!
日本語での自己紹介はこちらへ
Skills
Programming & Platforms
Linux,
MacOS,
Windows (+WSL),
UNIX command-line,
Python,
SQL,
R,
C++,
Bash,
Git,
GitHub/GitLab
Python
Numpy,
Pandas,
Polars,
Matplotlib,
Streamlit,
Pydantic,
Optuna
Machine Learning
TensorFlow,
Pytorch,
JAX,
Keras,
Numpyro,
Sklearn,
Kedro
Cloud & MLOps
AWS,
SSH (headless servers),
Docker,
Traefik,
MLFlow,
DVC,
BentoML,
FastAPI
Experience
Data Scientist & ML Engineer
Helicon Technologies, Sweden
Mar. 2024 — ...
- Developed a building optimization framework based on time series forecasting and gradient-free optimization.
- Designed and developed an in-house ML forecasting library toolbox.
- (Consultancy) Developed a real-time anomaly detection pipeline and pushed it to production on AWS (Kedro, MLFlow, Docker, BentoML). Responsible for model development and MLOps.
Senior Data Scientist
Twigeo, Sweden
Sep. 2022 — Feb. 2024
- Designed, trained, and deployed time series Bayesian machine learning models.
- Developed a scalable Bayesian Synthetic Control Method package for geo-testing (GeoLift AI); designed and analyzed geo-tests for many clients such as The New York Times and Duolingo.
- Built and deployed cloud-based WebApps, enabling scalability and accessibility for both technical and non-technical teams.
- Designed and maintained the data team MLOps and DeVOps platform (MLFlow, DVC, servers, …).
- Mentored and trained junior team members in advanced data science techniques.
Education
PhD in Physics
Uppsala University, Sweden
Aug. 2017 — Jul. 2022
- Published in high-quality journals such as Nature.
- Ph.D. representative.
- teacher assistant for the Quantum Physics laboratory class.
Master In Physical Engineering
Catholic University of Louvain, Belgium
Sep. 2015 — Jun. 2016
- Summa Cum Laude.
- Treasurer for the university-funded Japan-related project 'KotManga' & organizer of Japanese-French language exchange events (日本語話そう会).
- Erasmus Exchange Program in NTNU Trondheim, Norway (Sep. 2015 - Jun. 2016).
- Student representative.
Bachelor in Engineering
Catholic University of Louvain, Belgium
Sep. 2012 — Jun. 2015
Certificates
日本語能力試験 JLPT N2
August 2024
Issued by: The Japan Foundation & Japan Educational Exchanges and Services
February 2022
Issued by: IBM (Coursera)
December 2020
Issued by: University of Pennsylvania (Coursera)
March 2020
Issued by: DeepLearning.ai (Coursera)
August 2019
Issued by: University of Washington (Coursera)
Awards
-
Best Poster Award, European School on Magnetism 2018
September 2018 | Kraków, Poland
Best poster award for presenting my work on antiferromagnetic spintronics.
-
Bronze Medalist, International Physics Olympiad 2012
July 2012 | Tallinn, Estonia
Belgian representative at the International Physics Olympiad 2012 and Bronze medalist.
-
Excellence Prize
June 2012 | Collège Sainte-Marie, Saint-Ghislain, Belgium
Awarded for graduating as the top student of my year in high school, recognizing outstanding academic performance and overall excellence.