👨‍💻 Programming Languages

🧠 Machine Learning / AI / NLP

🌐 Web & Backend Development

⚙️ DevOps & Tools

🗃️ Databases & Data Engineering

📊 Visualization, Dashboards & UI

📚 Currently Exploring / Learning

🎓 Data & AI Engineer | Machine Learning Enthusiast | Software Developer

🚀 Passionate about AI, Data Science, and Software Engineering, I specialize in Machine Learning, NLP, CV, Deep Learning, and Backend Development. Whether it's developing ML-powered solutions, optimizing graph-based algorithms, or deploying scalable applications, I love solving complex problems with data-driven insights.

• 🎓 Currently: Master's student (Engineer's Degree, also known as Cycle d'Ingénieur in French) in my final year at ESILV, specializing in Data Science & AI.
• 🏢 Past Work: Experience in credit risk modeling, graph mining, and recommendation systems
• 🛠️ Tech Stack: Python, SQL, TensorFlow, PyTorch, Scikit-learn, Flask, FastAPI, Docker, CI/CD, Neo4j
• 📚 Projects: NLP-driven TripAdvisor Recommendation System, Real-time Chat Application, Graph-based School Donations Analysis, and more
• 🎯 Interests: AI for Cybersecurity, Finance, and Healthcare, Edge AI, and Scalable ML Systems

📌 🔍 CV–Job Offer Matching with LLMs (Internship Project @ Astek, 2025)

• Designed and evaluated hybrid AI systems for automated candidate-job matching
• Benchmarked open-source LLMs (Phi-3, Mistral, DeepSeek, Gemma) under zero-shot & few-shot settings
• Built pipelines for justification-aware classification, using label-only and label+justification prompting
• Integrated explainability using SHAP & ontology-based reasoning (GraphRAG-ready knowledge graph in OWL)
• Delivered a scientific-style evaluation report and contributed to an in-house R&D framework

📌 📊 Rating Prediction from Insurance Reviews (NLP Project)

• Built a Streamlit app to predict customer satisfaction ratings from insurance reviews
• Used RoBERTa, Neural Networks, and classical models (e.g., Random Forest, Logistic Regression)
• Applied SHAP for local interpretability and LDA for topic modeling

📌 📎 Graph-based School Donation Analytics

• Applied community detection (Louvain algorithm) on school-donor networks
• Built an interactive dashboard for visualizing patterns and donor affinities
• Extracted insights to help NGOs understand geographical and thematic funding gaps

📌 💬 Real-time Chat App (CI/CD + Docker)

• Architected a production-grade chat application using Flask, Docker, and GitHub Actions
• Integrated microservices, containerization, and real-time communication
• Scored 20/20 at ESILV for technical design and code quality

📌 🏨 TripAdvisor Recommendation System (NLP)

• Built a BM25-based baseline and enhanced with semantic similarity modeling
• Extracted latent features from user reviews and improved ranked retrieval performance
• Focused on unsupervised learning, without using direct labels

📌 🚧 Pothole Detection App (CV + Kotlin)

• Trained and exported YOLOv5, YOLOv6,YOLOv8 and YOLOv11 models to TFLite
• Built a real-time android app using Kotlin to detect potholes for road safety
• Benchmarked models on 665 labeled images, used data augmentation and explainability techniques

📌 📅 Smart Calendar Assistant (In Progress)

• Designing a web-based personal assistant that auto-schedules tasks based on user availability and travel time
• Uses RATP API, AI reasoning, and real-time Apple Calendar integration (Europe-focused MVP)

💡 Machine Learning & AI

• Supervised & Unsupervised Learning
• Explainable AI & Model Interpretability
• Deep Learning (CNNs, RNNs, Transformers)

💻 Development & Engineering

• Backend: Flask, FastAPI, Node.js
• Databases: SQL, MongoDB, Neo4j
• DevOps: Docker, CI/CD, GitHub Actions

📊 Data Science & Analytics

• Graph Mining & Network Analysis
• Natural Language Processing (NLP)
• Recommender Systems & Ranking Algorithms

🌐 Website/Portfolio: ahmedmaaloul.engineer
💼 LinkedIn: in/ahmedmaaloul
📧 Email: [email protected]

💡 Always open to collaborations, research, and new opportunities! 🚀