Sarah Zouinina

► Defined and implemented enterprise frameworks for the evaluation, design, and deployment of AI agents, reducing duplicated GenAI initiatives across teams. ► Established evaluation criteria for AI agents, covering performance, efficiency, ROI, and maintainability within an enterprise IT architecture context. ► Built and maintained an internal catalog of AI agents and tools to compare internal solutions and vendor products through structured Proofs of Concept. ► Benchmarked RAG-based agents leveraging Confluence data against vendor solutions (including Atlassian ROVO), assessing answer quality and user experience. ► Designed and tested “Talk to Your Financial” use cases using Text-to-SQL agents for automated access to financial data. ► Developed and evaluated agents using Dust, Databricks Agents, and Databricks Genie, applying advanced Prompt Engineering and RAG patterns. ► Collaborated with Enterprise Architecture and Internal IT teams to align GenAI agent initiatives with governance, scalability, and long-term platform strategy.

► Prepared and structured data pipelines for Retrieval-Augmented Generation (RAG), optimizing document ingestion and processing using Azure Document Intelligence and OCR techniques. ► Developed and enhanced search indexes with Azure AI Search, integrating indexers, skillsets, and data sources to improve information retrieval performance. ► Built and deployed GenAI solutions using Large Language Models (LLM), Prompt Engineering, and vectorization for intelligent document search and Q&A. ► Created automated CSV data extraction agents for large datasets, enabling processing beyond standard LLM context limits. ► Designed and implemented anomaly detection workflows for text-based datasets in geological estimation projects. ► Collaborated in architecture workshops and led CI/CD pipeline deployment to automate testing and production processes in an Agile/Scrum environment. ► Presented chatbot and GenAI solutions to stakeholders, delivered demos, and gathered feedback to drive continuous product improvement.

► Conception and creation of data ingestion pipelines from different sources (REST APIs, SQL Server, Postgre SQL ..) ► Data Transformation using Spark notebooks and/or Stored Procedures ► Migration from SAP BW to Synapse

► Worked on the ingestion and the tranformation of sales' data using Spark/ Scala, Git, Kubernetes, Jenkins and Airflow. ► Developed indicators about the quality of the affluence data pipeline using Python and Dataiku. ► Supervised projects on data cataloging and governance using DataGalaxy, Python, Azure Data Lake Service.

► Developed an ML‑based app for resumes and job‑offers matching using Transformers.

► Teached Computer Science and Mathematical foundations of Data Science: Web development with Database Integration, Linear algebra and Probability theory

► Extracted the text data and analyzed it using spectral clustering and dimensionality reduction techniques. ► Prepared reports on the behavioral patterns in the data streams and detected the hidden psycholinguistic phenomenons in collaboration with linguists and psychologists.

► Data anonymization using machine learning models state of the art.

► Provided daily dashboards with key performance indicators to evaluate the efficiency of marketing and commercial strategies. ► Identified the main revenue streams and conducted predictive analysis through Salesforce tool. ► Kept track of the company’s e-image, anticipated clients needs and market threats.

Simulation and Optimization of the Chlore Supply Chain ► Modeled the supply chain process using Arena solutions software, optimized the procurement system to cut cost by 28% and proposed the exact placement of the warehouses of Chlore to cut charges. ► Optimized an initial solution to the Vehicle routing problem what reduced the charges of procurement by 10%.