Bolzano, Italy
Mert Tezcan
Data & BI Consultant | Cloud Data Engineer
Fabric and Azure data platforms, with applied AI in the same repo — agents on live semantic models, document intelligence, and MCP into Fabric and Power BI.
About
Production data, applied AI.
I design data platforms that scale with the business — lakehouses, pipelines, and applied AI that takes unstructured work out of the loop.
I hold an MSc in Computational Data Science from the Free University of Bozen-Bolzano (unibz). At Würth IT Italy that means Microsoft Fabric, PySpark, and Power BI, plus an AI layer on top: MCP servers, Fabric Data Agents, and specialized agents that query live semantic models. Earlier work used Azure OpenAI and GPT Vision to turn messy documents into structured data and to automate classification and recommendations — models in the pipeline, not as a side demo.
-
MSc Computational Data Science
Free University of Bozen-Bolzano (unibz)
Bolzano, Italy -
BSc Computer Engineering
Istinye University
Istanbul, Türkiye
Experience
Log
-
Data & BI Consultant
activeWürth IT Italy · Bolzano, Italy · Jan 2026 – Present
Microsoft Fabric, Power BI, and applied AI on Würth’s BI landscape — lakehouses, semantic models, and agents that run against live data.
- Design lakehouse and analytics solutions on Microsoft Fabric: PySpark notebooks, metadata-driven pipelines into OneLake, Direct Lake semantic models, and Power BI.
- Build the AI layer on that stack: Fabric Data Agents, MCP (Fabric, Power BI, Dataverse), and agents that query live models via DAX instead of pasting extracts into a chat.
- Ship specialized BI agents — model readiness, platform governance, call analytics — plus orchestrator workflows that chain them.
- Drive agentic coding for Fabric and Power BI (VS Code, MCP, Claude via Azure) so platform work and AI work live in the same repo.
-
Cloud Data Engineer
Fruitservice · Bolzano, Italy · Jan 2024 – Jan 2026
Independent fresh-produce supplier and a long-standing commercial partner of Lidl since 1985. Azure data platform plus AI extraction on messy operational documents.
- Built and operated Azure data platforms with Synapse, Microsoft Fabric, and Power BI to serve internal teams and Lidl’s data needs.
- Led AI automation with Azure OpenAI GPT Vision: pull structured product data from unstructured supermarket PDFs, then land it in master data instead of leaving it in a chat window.
- Shipped containerized apps with Docker and Azure Container Apps, plus CI/CD for consistent build, test, and deploy.
- Automated manual processes with Azure Functions, Logic Apps, and Python ETL — end-to-end pipelines that unify retail data and cut operational overhead.
-
Data Scientist
Nubovi · Rotterdam, Netherlands · Jun 2023 – Jan 2024
Forecasting and LLM-backed insight on Azure cost and stakeholder decisions.
- Developed time-series models to forecast Azure spend and surface cost-optimization opportunities.
- Used GPT models for personalized recommendations and explanations stakeholders could actually follow.
-
Data Science Research Intern
Data Science Research Group, TalTech · Tallinn, Estonia · Aug 2021 – Oct 2021
Research on graph-based bot detection in social networks.
- Built models to detect fake-news-spreading Twitter bots with knowledge graphs and random forests (88% accuracy).
- Used Python, DGL, and Node2Vec to construct knowledge graphs and improve model performance.
Stack
Toolchain
Runtime
- Python
- SQL
- PySpark
- FastAPI
- Docker
- Git
- Azure
- Microsoft Fabric
- Power BI
- Databricks
- Synapse
- CI/CD
Applied AI
- Azure OpenAI
- GPT Vision
- LLMs
- MCP
- Claude
- Fabric Data Agents