Bechir Baaziz

Designed and implemented an AI-assisted system that generates React UI code based on user input (text prompts) using internal component libraries. Implemented a RAG (Retrieval-Augmented Generation) system with FAISS, to retrieve relevant UI components from internal libraries for accurate code generation. Built and tested a custom MCP (Model Context Protocol) NodeJs server to interface between client applications and LLMs like GPT and Claude via AWS Bedrock. Created Dockerized images for MCP and RAG, and deployed them on Kubernetes using Helm charts with service accounts and IAM roles for AWS integration. Created CI/CD pipelines using Jenkins for deployment. Built an interactive frontend interface using AnythingLLM, enabling users to chat with the system and visualize the generated UI output.

Contributed as part of a development team to the design and implementation of a full-stack web application for managing ADSL subscriptions and customer records, streamlining daily operations. Collaborated with technical and business teams to design and implement a REST API for internal service management. Worked in a collaborative environment, taking part in discussions, code reviews, and requirement analysis.