H
e
l
l
l
o
(
)
;
I
'
m
A
y
m
e
n
L
a
s
s
o
u
e
d
I
b
u
i
l
d
&
d
e
p
l
o
y
D
L
/
M
L
S
o
u
t
i
o
n
s
.
Data Scientist with a great experince in designing, building and deploying deep learning / machine learning solutions
Passionate about leveraging data-driven approaches to solve real-world problems and improve decision-making processes.
Hi! My name is Aymen. I'm a multidisciplinary engineering student with strong skills in data science and machine learning. Visit to learn more about me and my projects. I am always seeking real-world ai solutions and staying up-to-date with data science news.
I am interested in math, statistics, and computer science concepts, which is why I have chosen data science as a domain to develop myself in. I have been active on Kaggle daily for almost a year, learning and enjoying the world of data science, and following highly skilled individuals. This has greatly contributed to my improvement.
My main focus is continuous improvement. I am always looking to learn more and seeking new challenges.
Python
Java
C
Shell
HTML
CSS
JS
Neo4J
Supabase
MySQL
PostgreSQL
Pytorch
TensorFlow
OpenCV
ScikitLearn
Falsk
NodeJS
Laravel
Jupyter Notebook
HuggingFace
Chat GPT API
LangChain
Andoid Studio
Git
GitHub
Docker
Figma
A dynamic movie recommendation app created with Android Studio and Python. It utilizes web scraping, pretrained models, and a Flask server for personalized recommendations and Langchain for insightful insights.
The goal is to digitize all EPT services and enhance communication between teachers, administration, and current and former students, including graduates from various generations, to provide better guidance in the engineering curriculum. This project, similar to a Facebook clone, features a chatbot for an improved user experience and separate user and admin panels.
To support the Deaf community, I contributed as part of a team in training a model for real-time sound-to-ASL translation using ASL Gloss . I successfully trained the model on a privately curated dataset of videos, enabling it to translate recognized words directly into their corresponding signs, eliminating the need for letter-to-letter conversion.
After learning C, I wanted to conclude my journey with a project, which led to the creation of EPT Soduku Puzzler . Through this project, I explored new algorithms such as backtracking and enhanced my problem-solving skills.
I have developed numerous notebooks focused on object detection, deepfake image and video classification, and various applications in NLP and LLMs, including extensive fine-tuning with techniques like LoRa and QLoRa. My projects include teeth detection and numbering, creating RAG-based applications, as well as addressing diverse classification and regression problems through Kaggle and local competitions. I have also transformed several academic papers into code and engaged in data processing and training with large language models (LLMs) and natural language processing (NLP). My interests extend to accelerating deep learning processes using the Hugging Face accelerator and understanding the data flow within hardware components. All notebooks, competition solutions, and paper transformations are stored privately on GitHub for personal use.