Juan D.

I design and build data platforms in Microsoft Fabric from end to end: ingestion, Lakehouse, transformation, semantic model, and reporting in Power BI.

I work with companies that are at one of these three points:

They have data spread across Excel, SQL Server, SaaS, and on-premise, and need a single platform where the business trusts the numbers.

They are coming from Synapse, Databricks, or a legacy stack and want to modernize to Fabric without breaking what already works.

They have slow Power BI, duplicated models, or pipelines that break on Monday mornings, and need a clean architecture behind it.

How I work

I start by understanding the business case before touching the tool. I define the architecture (medallion: bronze / silver / gold), build the ingestion and transformation (Pipelines, Dataflows Gen2, Spark, SQL), and deliver a semantic model in Power BI on which the business can confidently exploit data. All code lives in Git, with tests and CI/CD when the project justifies it.

My background

10+ years in data: 7 at Thales (Engineering, Transport) and 3 at Millicom (telco), designing data models, optimizing Power BI, and leading migrations to Azure. Certified in Azure Data Engineer Associate (DP-203) and Microsoft Fabric Data Engineer (DP-700).

Types of projects that fit well

Design and implementation of a Fabric platform from scratch (4-12 weeks)

Migration from Synapse or other DWH to Fabric

Audit and redesign of Power BI models with performance issues

Design of the transformation layer (medallion + dbt or native Spark/SQL)

I can design a workflow with agents to increase the productivity of BI analysts and developers.

I can also design workflows with agents based on Hermes at a very low operational cost.

I also have experience building Generative AI applications on corporate data (RAG, conversational agents on BI models).