🎓 Computer Engineering Student @ ENICarthage
🚀 AI Engineer | Full-Stack Developer | MLOps Enthusiast

I build intelligent systems combining Artificial Intelligence, full-stack development, and DevOps automation. My expertise lies in creating scalable, production-ready applications that solve real-world problems.

• Natural Language Processing: RAG architectures, LLMs (LangChain, OpenAI), sentiment analysis, chatbots
• Computer Vision: CNNs, facial emotion recognition (BLIP-2, LLaVA), object detection
• Deep Learning: TensorFlow, Keras, PyTorch — classification, regression, clustering
• MLOps: Model deployment, monitoring, CI/CD for ML pipelines

• Backend: Spring Boot (Java), FastAPI (Python), Node.js — REST/SOAP/gRPC/GraphQL APIs
• Frontend: Angular, React, TypeScript — responsive, user-centric interfaces
• Architecture: Microservices, API Gateway, JWT authentication, workflow automation

• Containerization: Docker, Kubernetes — multi-service orchestration
• CI/CD: GitHub Actions, Jenkins — automated testing, deployment, monitoring
• Monitoring: Prometheus, Grafana — real-time performance tracking
• Security: SonarQube, OWASP ZAP — vulnerability detection, compliance

AI-powered HR assistant automating 200+ monthly queries with 60% time reduction.

Tech Stack: RAG architecture, LangChain, ChromaDB, FastAPI, Angular, PostgreSQL, OCR, JWT
Key Features:

• Multi-format document extraction (PDF, DOCX, PPTX, MSG)
• Semantic chunking & vector search (ChromaDB embeddings)
• LLM integration (Mixtral-8x7B) for contextual responses
• RBAC authentication, audit logging, GDPR-compliant

Compound facial expression recognition using BLIP-2 & LLaVA vision-language models.

Tech Stack: PyTorch, Transformers, Computer Vision, Streamlit
Key Features:

• Multimodal emotion detection (happy-sad, angry-surprised)
• Vision Transformer fine-tuning
• Real-time inference dashboard

Full-stack solution automating 300+ annual internships with 70% efficiency gain.

Tech Stack: Spring Boot, Angular, MySQL, Docker, JWT, OCR
Key Features:

• Multi-level validation workflow (Service → Committee → Direction)
• Intelligent PDF validation (OCR, stamp verification)
• Automated document generation (offer letters, certificates)
• Role-based dashboards for 800+ students & 5 actor types

Polyglot microservices architecture managing urban services (mobility, energy, emergencies).

Tech Stack: Spring Boot, Python (gRPC), Node.js (GraphQL), Docker, JWT
Key Features:

• 5+ independent microservices (REST, SOAP, gRPC, GraphQL)
• Database isolation (PostgreSQL, MySQL, MongoDB)
• Centralized API Gateway with JWT authentication
• Real-time monitoring dashboard (React)

98% accuracy CNN model with interactive drawing interface.

Tech Stack: TensorFlow, Keras, NumPy, Streamlit
Key Features: Real-time prediction, confidence scoring, data augmentation

Analyzed 1.6M tweets using TF-IDF, Logistic Regression & Naive Bayes.

Tech Stack: NLTK, scikit-learn, Pandas, Streamlit
Key Features: Real-time sentiment prediction, confidence scoring, comparative model analysis

✓ 60% reduction in HR processing time (AskRH chatbot)
✓ 70% efficiency gain in internship management (300+ workflows automated)
✓ 98% accuracy in CNN-based digit recognition
✓ 1.6M tweets analyzed for real-time sentiment prediction
✓ 5+ production-ready applications deployed with containerization

💼 Open to: AI/ML Engineer | Full-Stack Developer | MLOps Engineer roles
🌍 Languages: Arabic (Native) | French (B2) | English (B2)

⭐️ From Mahdi-toumi