Moataz Bouazizi

– Architected and built end-to-end a multi-agent Generative Al SaaS platform for talent intelligence, enabling companies to analyze candidate profiles and optimize hiring decisions. – Designed and implemented a Retrieval-Augmented Generation (RAG) pipeline, including the retrieval layer and embeddings, to ground candidate evaluation in CVs, job descriptions, and interview data. – Orchestrated multiple coordinating Al agents combining retrieval, prompt engineering, and structured reasoning to automate end-to-end candidate evaluation. – Deployed the platform to production, used in real evaluations by recruiters and clients, and implemented monitoring and evaluation metrics to track and improve output quality.

– Architected and built end-to-end a multi-agent GenAl system to generate, evaluate, and refine supply chain optimization scenarios based on business rules and operational constraints. – Designed and implemented data acquisition and preprocessing pipelines for heterogeneous operational data, ensuring quality and temporal consistency for downstream Al workflows. – Deployed the solution on Azure with containerization and CI/CD, industrializing the system to production. – Defined KPI-based evaluation mechanisms to validate generated scenarios against business rules, driving continuous improvement of agent outputs.

– Developed an Al-powered infrastructure intelligence platform for automated analysis of operational data, enabling anomaly detection, knowledge extraction, and decision support. – Built end-to-end data pipelines for processing heterogeneous data sources, including data preprocessing, transformation, semantic analysis, and structured information extraction. – Designed Generative Al workflows using LLM-based approaches, prompt engineering, embeddings, and retrieval techniques for contextual information retrieval and automated insights generation. – Implemented Python-based Al services and APIs to integrate machine learning and GenAl capabilities into scalable enterprise applications. – Applied model evaluation, Docker containerization, and MLOps practices to improve Al system reliability and deployment readiness.