Mohamed El Haddioui

Delivered 10+ AI proof-of-concept projects for enterprise clients, gathering requirements, defining acceptance criteria, and delivering production-ready demonstrators in 2-4 week sprints. Designed and implemented RAG systems using LangChain, LlamaIndex, and hybrid retrieval to improve retrieval accuracy and reduce hallucinations in knowledge-intensive applications. Integrated modern LLMs (DeepSeek, Qwen, OpenAI) into business workflows via prompt engineering, conversational memory, and tool-augmented agents to automate internal decision-making tasks. Maintained high software quality with Git workflows, comprehensive API documentation (OpenAPI/Swagger), and automated testing using pytest and unittest.

Delivered 10+ AI proof-of-concept projects by collaborating with enterprise stakeholders to gather requirements, define acceptance criteria, and deliver production-ready demonstrators in 2-4 week sprints. Designed and implemented RAG systems using LangChain, LlamaIndex, and hybrid retrieval, significantly improving retrieval accuracy and reducing hallucinations in knowledge-intensive applications. Integrated modern LLMs into business workflows via prompt engineering, conversational memory, and tool-augmented agents that automate internal decision-making tasks. Ensured high software quality with clean Git workflows, comprehensive API documentation, and automated testing using pytest and unittest.

Developed and deployed a Laboratory Information Management System (LIMS) using event-driven microservices architecture. Built multiple Spring Boot microservices orchestrated via Spring Cloud Gateway and service discovery, utilizing Apache Kafka for asynchronous messaging and high service availability. Secured REST APIs with Keycloak, implemented role-based access control, and documented all endpoints using OpenAPI 3.0 for seamless integration with client applications.