Mohamed Fedi LETAIEF

• Reduced report turnaround time by automating X-ray analysis and preliminary report generation with doctor-in-the-loop validation. • Trained and deployed YOLO models for dental imaging (teeth detection & anomaly classification). • Built a doctor-in-the-loop active-learning pipeline via Label Studio to continuously improve model performance. • Developed a LangChain-based VLM pipeline for treatment suggestions combining CV outputs with rule-based logic. • Built a FastAPI inference service integrated with Odoo for real-time AI-driven clinical and sales workflows. • Designed and developed a secure patient portal using Next.js (BFF architecture) enabling patients to access appointments, treatments, documents, and AI-assisted messaging. • Engineered an MCP server exposing Odoo as LLM-callable tools, enabling natural language CRM operations. • Improved CRM adoption and sales efficiency by enabling non-technical users to manage workflows via AI. • Containerized and deployed the system with Docker for scalable, reproducible environments

• Designed and built scalable RESTful APIs using Hono.js and TypeScript; optimized PostgreSQL schemas, indexing, and triggers. • Integrated Stripe payments, Cloudflare R2 S3 storage, OAuth/JWT authentication, and role-based access control. • Embedded OpenAI API with prompt engineering to automate workflows and enhance user experience. • Streamlined CI/CD pipelines (GitLab) and containerized deployments with Docker.

• Led the design, development, and maintenance of NoteShare, an AI-powered platform for students to share, access, and collaborate on academic documents. • Implemented robust user management systems, including registration, authentication, and authorization features to secure user data and platform functionalities. • Integrated OCR technology for automatic text extraction and duplication detection, enhancing search functionality and content management. • Developed and managed relational databases. • Built a real-time chat system, enabling communication between students.