// AI ENGINEER & RESEARCHER

MAJD SOUISSI

Engineering Student in Artificial Intelligence | Data Science Researcher

Explore My Work 📄 Download CV

About Me

Majd Souissi

I'm an Engineering student at the National Higher Engineering School of Tunis (ENSIT), specializing in Applied Mathematics and Modeling with a deep passion for Artificial Intelligence and Data Science.

Currently pursuing a Research Master in Data Science (M2), I focus on cutting-edge AI research including Vision-Language Models, Explainable AI, and multimodal systems. My work spans computer vision, NLP, and deep learning applications.

With hands-on experience in AI research internships and multiple published papers, I'm dedicated to pushing the boundaries of what's possible with AI while ensuring interpretability and real-world impact.

10+
AI Projects
15+
Certifications
930
TOEIC Score

Technical Arsenal

// AI & Deep Learning

PyTorch TensorFlow Keras Hugging Face LangChain NVIDIA NeMo/NIM OpenCV U-Net ViT CLIP Qwen-VL LLMs RAG BERT LSTM

// Explainable AI

LIME SHAP Grad-CAM XAI Techniques Model Interpretability

// Development & MLOps

Python C/C++ Java (JEE) Dart (Flutter) Flask FastAPI Docker ONNX Git/GitHub CI/CD

// Data Science & Analytics

Scikit-learn XGBoost Power BI Tableau IBM SPSS Seaborn Matplotlib Pandas NumPy

// Databases & Big Data

SQL PL/SQL MySQL MongoDB Oracle DB FAISS Chroma Kafka

// Mathematics & Optimization

Linear Algebra Probability Stochastic Processes Time Series (ARIMA, Prophet) Genetic Algorithms Linear Programming Operations Research

Professional Journey

AI Research Intern

Data Engineering and Semantic Research Unit, Faculty of Sciences Sfax

August 2025 – September 2025

  • Studied and implemented state-of-the-art Vision-Language Models for video captioning and semantic retrieval
  • Evaluated multimodal retrieval pipeline, testing models such as Qwen-VL 2.5 and InternVL
  • Conducted performance analysis and wrote detailed research report with comparative results
  • Keywords: Multimodal retrieval, Vision-Language, RAG, FAISS, Kafka, Python

AI Research Intern

Larina Research Laboratory for Intelligent Networks and Nanotechnology

June 2025 – August 2025

  • Developed explainable CNN for almond shell classification, interpretable via Grad-CAM and LIME
  • Integrated model into Flutter mobile application for real-time classification
  • Publication: Paper accepted at OCTA'2025 Multi-Conference - "Improving Almond Variety Identification through Wavelet-Based Image Augmentation"
  • Keywords: XAI, CNN, Grad-CAM, LIME, Flutter

Marketing Manager

NERDATA ENSIT CLUB

2023 – Present

  • Organized cutting-edge data science and AI workshops
  • Spearheaded promotion of club activities through strategic media channels
  • Coordinated comprehensive technical training for 50+ club members

Media Manager

VANGUARD CHESS ENSIT CLUB

2023 – 2024

  • Organized soft skills workshops and competitive chess tournaments
  • Led promotional activities through strategic media campaigns
  • Served as media and logistics manager for ENSIT GAMBIT 1.0

Logistics Manager

ENSIT Geeks Club Hackathons

2023 – 2024

  • Ensured seamless competition atmosphere for all participants
  • Coordinated complex logistics and technical requirements for multiple high-stakes hackathon events

Featured Projects

PneumoXAI: Interpretable Chest X-ray Analysis

Designed a CNN for pneumonia detection on chest X-rays with interpretability via Grad-CAM and LIME. Developed data augmentation pipeline including SMOTE for class balancing. Created Flask web application with LangChain chatbot for medical predictions and interpretations.

CNN Grad-CAM LIME SMOTE Flask LangChain XAI

PodSum: Intelligent Podcast Summarization

Developed pipeline combining Whisper for audio transcription and Qwen LLM for summary generation. Implemented RAG module enabling user-content interaction. Evaluated pipeline's performance in terms of accuracy and relevance on multiple datasets.

Whisper BERT Qwen LLM NLP RAG LangChain FAISS

AutoCoverLetterRAG – Automated Motivation Letter Generator

Built full-stack React.js + Node.js web application that generates motivation letters from user's CV and job description. Designed RAG pipeline combining document parsing, embeddings, and retrieval to link candidate skills to role requirements. Automated LaTeX compilation to produce clean, professional PDF motivation letters.

React.js Node.js LangChain RAG FAISS LLMs LaTeX

MindMate AI Chess Game

Built innovative web-based chess game powered by reinforcement learning, enabling AI to continuously learn and improve with experience. Features custom chess engine integrated with Flask backend and dynamic JavaScript interface.

Reinforcement Learning Chess Engine Flask JavaScript HTML/CSS

CryptoCrack AI

Developed AI-powered web application for encryption and decryption using classical cryptographic methods, enhanced with integrated chatbot assistance for seamless user interaction and educational support.

AI Decryption Cryptography Chatbot NLP Flask

Vision-Language Models for Video Captioning

Research project implementing state-of-the-art VLMs for video captioning and semantic retrieval. Evaluated multimodal retrieval pipeline with models like Qwen-VL 2.5 and InternVL, conducting comprehensive performance analysis.

Vision-Language Models Qwen-VL InternVL RAG FAISS Kafka

Certifications & Achievements

🏆

AWS Machine Learning Foundations

Amazon Web Services

📊

Data Scientist Associate

DataCamp

🤖

Deep Learning Fundamentals

NVIDIA

👁️

Computer Vision for Industrial Inspection

NVIDIA

🎨

Generative AI with Diffusion Models

NVIDIA

⚡

Rapid Application Development With LLMs

NVIDIA

🤝

AI Agents Fundamentals

HuggingFace

☁️

Azure AI Fundamentals (AI-900)

Microsoft

💾

Azure Data Fundamentals (DP-900)

Microsoft

Let's Connect

📧

Email

majdsouissi625@gmail.com

📱

Phone

+216-94986601

📍

Location

Tunis, Tunisia

💼

LinkedIn

majd-souissi

💻

GitHub

majdsouissi

🎓

Education

National Higher Engineering School of Tunis