Ferjani karim

I am an AI engineer and full-stack developer with hands-on production experience building intelligent web applications from the ground up. At Asteelflash, I designed and delivered a complete RFQ management platform using Django REST Framework, PostgreSQL, and React, implementing role-based access control, versioning, and structured approval workflows. At Telnet, I fine-tuned a BERT model that reached 92% accuracy on sentiment analysis tasks, serving real-time predictions through a Flask API with response times under 300ms. On the computer vision side, I trained a Mask R-CNN model on over 5,000 satellite images achieving 92% detection accuracy, applying transfer learning to cut training time by 30%. My core stack is Python, Django, Flask, React, PyTorch, TensorFlow, and PostgreSQL, with working knowledge of Docker, AWS, and Azure. I hold a Microsoft Azure AZ-900 certification and an NLP certification from Coursera. I work fluently in French, English, and Arabic, which makes me a natural fit for European and MENA-based clients, and I am available immediately from UTC+1.