Available for Work

Atef Bouzid

Data Scientist & AI Engineer

Transforming data into intelligent solutions through machine learning and AI. Building production-ready systems that solve complex problems.

View My Work Get In Touch Download Resume - CV
Last updated: Loading...
Atef Bouzid
Languages:
English French Arabic

Experience

Upcoming (Confidential for now)

04/2026 – 09/2026
Paris, France

Data Scientist

Upcoming

VINCI

01/2026 – 03/2026
Paris, France

Data Scientist - Capstone Project

  • Built an end-to-end AI pipeline for automated country risk reports (LLMs + RAG) — from web scraping and data processing to retrieval, generation, and analyst validation via a Streamlit app — cutting manual research and writing effort significantly.
  • Deployed on Azure (Azure OpenAI, Blob Storage, Cosmos DB) with secure REST APIs, delivering a scalable, production-ready solution.

ReDX Technologies

06/2024 – 07/2025
Remote & Sfax, Tunisia

AI Engineer

02/2025 – 07/2025
  • Designed and benchmarked a fully open-source recommendation system with LLM-based feature extraction using Qwen2.5.
  • Implemented multi-factor scoring algorithms and geospatial optimization for task-worker matching.
  • Collaborated with cross-functional teams to validate and benchmark product performance.

GenAI Intern

06/2024 – 08/2024
  • Developed an AI chatbot and recommender system using LLMs and RAG architecture, capable of processing diverse data formats (Excel, CSV, PDF) to automate HPC cluster configuration.

Hutchinson

07/2023 – 08/2023
Sousse, Tunisia

Computer Vision Intern

  • Developed a data pipeline and anomaly detection model achieving 97% accuracy for quality control, with statistical analysis, and deployed it on a user-friendly HMI interface.

Education

École Polytechnique - Institut Polytechnique de Paris

MSc (Master 2) Data Science: Statistics and Applied Math

09/2025 – 10/2026 Paris, France

Télécom Paris

MSc in Engineering - Specialization in AI and Image Processing

09/2024 – 06/2025 Paris, France

École Polytechnique de Tunisie

National Engineering Degree in Multidisciplinary Engineering

09/2022 – 06/2024 Tunis, Tunisia

Preparatory Institute for Engineering Studies of Monastir

Preparatory Studies

09/2020 – 06/2022 Monastir, Tunisia

Skills & Technologies

Programming Languages

Python 🐍
Python
SQL 🗄️
SQL
HTML 🌐
HTML
CSS 🎨
CSS

API Development & Databases

FastAPI ⚡
FastAPI
PostgreSQL 🐘
PostgreSQL
MySQL 🗄️
MySQL

DevOps & Cloud

AWS ☁️
AWS
Azure 🔵
Azure
Docker 🐳
Docker
Git 📦
Git
Apache Kafka 📨
Apache Kafka

ML/AI Technologies

PyTorch 🔥
PyTorch
TensorFlow 🧠
TensorFlow
Scikit-learn 📊
Scikit-learn
LangChain 🔗
LangChain
LangGraph logo 📈
LangGraph
Hugging Face 🤗
Hugging Face
OpenCV 👁️
OpenCV
Pandas 🐼
Pandas
NumPy 🔢
NumPy
Seaborn 📉
Seaborn
Matplotlib 📊
Matplotlib
Plotly 📈
Plotly
PyLab 🔬
PyLab
Dataiku 🐦
Dataiku

Projects & Achievements

🏆

Competitions & AI hackathons Won

8

🎓

Government Scholarship

Tunisia

Projects

A selection of my personal projects in Data Science and AI. Explore my open-source contributions and technical experiments on my GitHub profile.

AI Investment Research - InvestiAgent
InvestiAgent

AI Investment Research - InvestiAgent

Created full-stack AI platform generating stock and financial analysis reports (LaTeX/PDF) using LLMs (DeepSeek) and real-time data pipelines, reducing analysis time to 3 minutes through statistical modeling and time series analysis.

Python LLM Full-Stack Time Series
View Demo View on GitHub
Machine Failure Prediction, MLOps Dashboard
MLOps Dashboard

Machine Failure Prediction, MLOps Dashboard

Developed a machine learning system predicting industrial equipment failures with 96%+ accuracy using ensemble models. Built a FastAPI backend and Next.js dashboard for real-time monitoring, maintenance alerts, and production analytics.

Machine Learning FastAPI Next.js MLOps
View Demo View on GitHub
LabelSense: AI-Powered Textile Defect Detection System
LabelSense

LabelSense: AI-Powered Textile Defect Detection System

Developed an intelligent textile defect detection system using YOLOv8, achieving 94.2% precision and 91.8% recall for multi-class classification of printing errors, stains, and structural flaws, deployed as a real-time Flask web application.

YOLOv8 Computer Vision Flask Deep Learning
View Demo View on GitHub

Get In Touch

I'm always open to discussing data science projects, AI research collaborations, or opportunities to solve challenging problems with machine learning. Feel free to reach out!