I'm Mohamed Amine Karmous
AI & Software Engineer
Passionate AI Engineer dedicated to building intelligent solutions that drive innovation. Specializing in machine learning, natural language processing, and full-stack AI development. Let's transform ideas into impactful applications.
About Me
( 01 )
AI & Software Engineer specializing in LLM, NLP and its Arabic applications, LLM fine-tuning, chatbots, Machine learning, deep learning and computer vision. Experienced with Python, PyTorch/TensorFlow, Hugging Face, FastAPI, LangChain/LangGraph, n8n, and cloud deployment (Azure/AWS/GCP). Skilled in building scalable AI models and APIs for real-world applications with expertise in MLOps, microservices architectures, and production-ready deployment.
03
Years of experience
100+
Happy Clients
100+
Project Completed
Language
English
French
Arabic
Spanish
Experience
( 02 )
11/2025 – Present
AI & Robotics Engineer
Fulltime
Built and deployed enterprise AI systems using LangChain, OpenAI, RAG, and Pinecone. Developed advanced autonomous intelligence for Promobot V3. Designed large-scale sales and footfall forecasting models for Qatar Museums. Built AI-driven automotive analytics platform for Al Manaï dealership.
01/2025 – 10/2025
Full-Stack Generative AI Engineer
Fulltime
Built medical fertility-assistant chatbot using React, Spring Boot, FastAPI, and OpenAI/Claude models. Designed medical entity extraction with SciSpaCy and spaCy. Developed LangGraph multi-agent flows and microservices with Azure DevOps CI/CD pipelines.
07/2024 – 12/2024
AI Full Stack Developer
Fulltime
Built full-stack Angular + Spring Boot platform for EV license plate detection. Developed PyTorch-based segmentation model with 95% accuracy for Yolov8. Created ML regression model for EV consumption prediction with 91% accuracy. Added MLOps with MLflow and Kubernetes deployment.
03/2024 – 07/2024
AI and Computer Vision Developer
Fulltime
Built table detection and recognition system using YOLOv8 and PaddleOCR achieving 98% precision. Contributed to advanced OCR model evaluation and dataset annotation pipeline. Integrated MLOps workflow using Kubeflow and MLflow.
06/2023 – 09/2023
Mobile and AI Application Developer
Fulltime
Built complete delivery app with Flutter + Firebase. Designed hybrid recommendation system using TF-IDF, collaborative filtering, Word2Vec/FastText embeddings.
Education & Skills
( 03 )
Artificial Intelligence Engineering Degree - 2025
Specialized in AI, Machine Learning, and Deep Learning from Epi Digital School
Bachelor's Degree in Computer Science - 2022
Graduated from Isitcom with focus on software development and computer science fundamentals
Python
PyTorch
TensorFlow
LangChain
FastAPI
Kubernetes
Docker
Azure
OpenAI API
React
Next.js
Spring Boot
Latest Works
( 04 )
multi agent radiology assistant
Client: qatar hospitals
Multi-agentic ai assistant for hotel bookings and recommendation
Client: St regis doha
Almanai-group-heavy equipements sales forecasting and recommendations
Client: almanai-group
Amara Hospitality booking agent
Client: Katara-hospitality
AI trip planner multi-agent
Client: freelancing project
AI-Powered Trip Planner built with LangGraph, LangChain, OpenAI, and agent-based workflows to create intelligent and personalized travel plans. The system uses multiple AI agents and specialized tools to handle different travel-planning tasks, including flight and airport information, ticketing, weather forecasting, currency conversion, and destination planning. The LangGraph workflow orchestrates the agents and tools to process user requirements, retrieve real-time travel information through external APIs, and generate optimized itineraries. The application leverages the OpenAI API for natural-language understanding and decision-making, creating an interactive and automated travel-planning experience.
AI multi agent cv recruitement system
Client: freelancing project
**AI-Powered CV Recruitment System** built with **LangGraph, LangChain, OpenAI, and Streamlit**. The system uses a multi-agent workflow to automatically parse and analyze candidate CVs, evaluate **skills, professional experience, and cultural fit**, calculate an aggregate candidate score, and generate a final **hiring recommendation**. It provides an interactive Streamlit interface with detailed analysis dashboards and supports **JSON report generation, download, and local assessment persistence**. The system is designed to streamline and automate CV screening and recruitment decision-making using Generative AI and agentic workflows.
Cats Breed prediction
Client: personal project
Cars number plate detection and recognition
Client: Horizop energy
AI-Powered License Plate Recognition & EV Platform combining YOLOv8, EasyOCR, CNN, Selenium, BeautifulSoup, and web scraping technologies. The system detects and recognizes vehicle license plates from images and video, with specialized support for Tunisian Arabic license plates using a custom CNN model to improve recognition accuracy. It also integrates automated web scraping with Selenium and BeautifulSoup to retrieve vehicle information associated with detected license plates from online sources across the Netherlands, United Kingdom, France, and Denmark. In addition, the project includes an Electric Vehicle (EV) management platform that enables users to locate and book charging stations, manage charging sessions, and facilitate the charging process for electric vehicles. The solution combines Computer Vision, OCR, Deep Learning, Web Automation, and EV management into an integrated platform.
diabet disease prediction
Client: Freelancing project
Multi-agent-ai fertility specialised assistant
Client: Tunsian and north african clinics
Toxic comment detection application
Client: Smart conseil company project