Leonard Cseres
Data Scientist
Web Developer
Designer
About
My Story
As a data-driven entrepreneur with a passion for tech, I bring a unique combination of skills to solve human problems. With my knowledge in UI/UX design, data analytics, software and web development combined with business mindset, I bring to life unique ideas.
Ecole Polytechnique Fédérale de Lausanne
My time at EPFL was instrumental in taking the right development path. My passion for computer science was ignited here.
2020-2023
EPFL Entrepeneur Club
During my studies, I was a member of the EPFL Entrepreneur Club where I was able to learn from and collaborate with other entrepreneurs.
2021-2023
Scrapeout
With interest in data analytics, I founded Scrapeout. I was able to learn about real life entrepreneurship and manage challenges along the business growth.
2021-2023
HEIG-VD
I am currently studying at HEIG-VD in order to deepen my knowledge in software development. I am also working as a student assistant working on various ML and MLOps projects.
2023-nowMy Favorite Works
Works
Greenland Supraglacial Lakes Detection
This project was part of a ACM SIGSPATIAL competition. The goal was to segment supraglacial lakes in Greenland using satellite images. We worked in a team of 5 at HEIG-VD and submitted two solutions, ranked 1st and 3rd.
2023 ACM
SIGSPATIAL
1st Place
2023 ACM
SIGSPATIAL
3rd Place
Works
Swiss License Plate Detection
This project was part of research into Machine Learning Operations (MLOps) at HEIG-VD. It is comprised of two models: one for detecting the license plate and another for recognizing the characters. The key feature is a custom license plate generator in order to train on synthetic data.
Works
Scrapeout
Scrapeout is my most ambitious project to date. It has now become a full web tool that helps companies identify conflict of interest in their data in order to take more informed decisions.
Hackathon
1st Place
AWS
Certified
Works
Captcha Solver
The goal of this project was to create a neural network that could solve 6 character captchas from Amazon. I achieved an average accuracy of 93%.
Works
SlitherIO RL
The project was to create a neural network that could play the game SlitherIO. The agent was trained using A2C (actor to critic) reinforcement learning. The agent was able to collect food and avoid other snakes.
Works
Pac-Man
Remake of the classic Pacman game. This project was done in a team of 2 for the Introduction to Programming (CS-107) course at EPFL.
It includes all ghosts and their AI, a map editor, a leaderboard system, custom menu and more.
1st Place EPFL
Award
Contact
Send me an Email
leonard.cseres@devleo.ch