Zakaria Moufannine

Context:WTW is a global leader in risk and insurance consulting (49,000 employees, 140 countries, ~$10 billion in revenue). As part of a data modernization program, two divisions were entrusted to me: Corporate Risk & Broking (CRB) and Health & Benefits (H&B), with different approaches based on business needs.

Before migrating, order had to be restored. The legacy Oracle data model carried years of accumulation: poorly defined entities, implicit business rules, undocumented dependencies.

I took charge of the complete redesign of the data model in direct collaboration with actuaries, brokers, and data providers. Once the model was stabilized, I led the migration to Azure Synapse Analytics and rebuilt the Power BI reporting to replace the old OBIEE reports.

Different approach: iso migration from Oracle to Microsoft Fabric. The goal was not to remodel, but to faithfully transpose existing objects and pipelines to the new Fabric/OneLake platform, ensuring GDPR compliance on sensitive health data.

The Power BI reporting was rebuilt identically, then optimized to leverage Fabric's native capabilities (Direct Lake, centralized semantic models).

Stack: Oracle · PL/SQL · Azure Synapse Analytics · Microsoft Fabric · OneLake · Power BI · DAX · OBIEE · Azure Data Factory · SQL · Azure DevOps

Technip Energies is a world leader in engineering for the energy transition (LNG, hydrogen, CO2). Since 2022, the group has launched an ambitious data transformation focused on "Data Products" — reusable data solutions from one project to another — driven by the group's CDO.

My mission: design and deliver, from scratch, the first ServiceNow data pipeline to the internal data platform — with Power BI dashboards to manage IT incidents.

What I built:

• ServiceNow API Ingestion: connection to ServiceNow's REST Table API, extraction of incidents, tickets, SLAs, and metadata. Pagination management (sysparm_limit / sysparm_offset) to process large volumes of data without loss or timeouts.

• Data Model: design of the analytical schema in Azure Synapse — fact tables (incidents) and dimensions (priority, category, group, status, SLA) — optimized for Power BI querying.

• Power BI Dashboards: delivery of IT incident management dashboards (volume, SLA, MTTR, distribution by team and category), providing management with reliable and up-to-date data for the first time.

Added value: a 100% greenfield project that replaced manual Excel extractions and gave management data-driven visibility into incidents — in line with the group's data transformation strategy.

Stack: ServiceNow REST API · Azure Synapse Analytics · SQL · Power BI · DAX · Python / Power Query

Context:Michelin is the world's second-largest tire manufacturer, committed to an ambitious Industry 4.0 transformation since 2019. One of the key axes of this digital roadmap: assisted access to information for factory operators. The on-site observation was clear — technicians and workers were wasting valuable time manually searching SharePoint for technical procedures, machine data sheets, and quality standards they needed in real-time.

My mission:design and deliver an RAG (Retrieval-Augmented Generation) solution on Azure to transform this document SharePoint into an intelligent conversational assistant.

• Intelligent SharePoint Indexing: connection via Microsoft Graph API, extraction of all technical documents in French (machine data sheets, procedures, ISO quality standards), chunking, embedding generation via Azure OpenAI, and vector indexing in Azure AI Search with hybrid search (semantic + keyword).

• Complete RAG Pipeline: for each question asked by an operator, the system finds the most relevant document passages via semantic search, injects them as context into the GPT prompt, and generates a precise, sourced answer — in French, in a few seconds.

• Operator Chatbot: conversational interface deployed for factory teams. The operator types a few words ("belt change procedure", "tire X quality control standard") and gets an answer extracted from official Michelin documents, with source citation.

Value delivered:zero manual SharePoint navigation, reliable answers anchored in internal documentation, and direct operational time savings for production teams — fully aligned with Michelin's "assisted access to information" vision.

Stack: Azure OpenAI · Azure AI Search · Microsoft Graph API · SharePoint Online · Python · RAG · Embeddings · Chatbot