Banu Vinay / virtualbuddy.in

Fabric Data Engineer & Trainer, Bengaluru

I build lakehouses and pipelines on Microsoft Fabric, and the governance that keeps them trustworthy.

Data analyst by origin, data engineer by practice. I build end-to-end Fabric platforms: metadata-driven ingestion, medallion lakehouses, star-schema warehouses, semantic models and the Power BI on top. I train the teams who inherit them, and I'm bringing Microsoft Purview into every build so the data stays cataloged, classified and traceable.

See my work Get in touch
A medallion data pipeline: sources flow into bronze, silver and gold lakehouse layers, then into Power BI, with Purview governing every stage. Microsoft Purview: catalog, lineage, classification, sensitivity labels ERP & CRM databases SaaS APIs & Delta Sharing SharePoint, files, events Bronze raw, as landed Silver cleaned, conformed Gold modeled for BI Power BI semantic model & reports Fabric Lakehouse: Pipelines, Dataflows Gen2, Spark notebooks, Delta tables
The shape most of my work takes. Sources land in bronze untouched, get cleaned and conformed in silver, and are modeled in gold for reporting. Purview sits across all of it.

Work

Selected projects. Each one lists the problem, what I built, and what changed. More detail on request.

2026 – present · Fire protection company, USAFabric F16 · Lakehouse · Warehouse · Pipelines · PySpark · T-SQL · Power BI

Enterprise data platform on Microsoft Fabric, built from zero

Problem
A fire-protection company with data spread across an Odoo ERP, Salesforce, Procore, InspectPoint, ADP and SharePoint, an IT team new to Fabric, and reporting that lived in one-off dataflows nobody could maintain.
Built
Sole data engineer embedded with IT leadership. Config-driven, metadata-driven ingestion into a medallion lakehouse (six source systems; PostgreSQL read replicas, the Salesforce connector, Delta Sharing, REST APIs and file drops), incremental loads with watermarks and MERGE patterns, a gold-layer star schema in a Fabric Warehouse refreshed by stored procedures, DirectQuery and Direct Lake semantic models, and an operational layer of config tables, run logging and a threshold-based data-quality framework. Then retired the legacy reports into governed gold models: unified time entries across ERP and field-ops systems with timecard-lateness analytics, and sales commissions.
Solved
PostgreSQL WAL conflicts on the read replica (retries, chunked loads), unscaled numerics breaking Spark casts (dynamic precision), and Fabric's own constraints: no nested procs, no ALTER TABLE, DirectQuery and Direct Lake limits.
Outcome
Production platform in a few months, plus a multi-session Fabric and Power BI training program delivered to the 8-to-10-person IT team.
2021 – 2026 · Fairgrove Property Management, USAAzure Functions · ADLS · Data Factory · Azure SQL · Power BI · Power Automate

Property and finance pipeline from AppFolio, end to end

Problem
A real-estate firm with property, finance and payroll data spread across AppFolio, Paylocity and SharePoint, and no single owner of its reporting.
Built
Sole analytics lead for four years. AppFolio APIs pulled by Azure Functions into ADLS, transformed with Data Factory, loaded to Azure SQL, served in Power BI. Automated PDF generation for tenant and client updates.
Outcome
Owned the firm's reporting data estate: stood up the pipelines and delivered Power BI reporting across operations, finance and leadership.
2020 – 2026 · Automation Strategy & Performance, USATalentLMS API · Azure Functions · Azure SQL · Data Factory · Power BI

Analytics foundation for a growing edtech business

Problem
Joined as one of the first four employees and stayed nearly six years. Reporting was manual, error-prone and already causing client-facing mistakes.
Built
End-to-end flows from TalentLMS, QuickBooks, HubSpot and Google Sheets into Azure SQL and Power BI, with monitoring and fallback logic for failed refreshes. Later rebuilt on Fabric: API to notebooks to bronze and silver Delta tables to Direct Lake models.
Outcome
Owned the reporting data estate end to end: pipelines, the SQL layer and Power BI reports for operations, finance, sales and client delivery.
2022 – 2024 · GroupM, USAPower BI · Excel · Power Automate

Global workforce analytics for 40,000 people

Problem
The Global HR Director needed headcount, attrition and diversity reporting across regions from people data that arrived in every shape imaginable.
Built
Collection, aggregation and Power BI dashboards for a 40,000-person workforce, with Power Automate and macros to speed the monthly cycle.
Outcome
Faster monthly executive reporting and first-hand exposure to enterprise data governance in a global matrix organisation.
Demos & side projectsFabric · SQL Database · REST APIs · GitHub Actions

Smaller builds worth a look

Write-back
From read-only to write-back: a native Power BI write-back into a Fabric SQL database using Translytical Task Flows, prompted by a hotel client who wanted to act from inside the report.
FinOps
Fabric Capacity Controller: a stateless scheduler built on the Fabric Capacity REST API, Bash and GitHub Actions that pauses and resumes capacity around workload windows, after I burnt my own Azure bill leaving one running.
Research
Architecture and live demo of Fabric Data Agents, conversational BI and Real-Time Intelligence (EventStream to EventHouse) for a US client, including the governance implications.

Feedback

From client reviews on Upwork. Full history on my profile.

A dedicated and reliable Azure DevOps and Fabric Data Engineer who consistently delivered high-quality work. She learns quickly, communicates clearly, and approaches every task with a strong work ethic. I would be happy to work with her again.

Fabric platform build · 5.0 · 2026

A true professional. Very trustworthy, with exceptional problem solving skills. Her profile does not tell her whole story. Very knowledgeable in cybersecurity, networking, Microsoft Azure and IT services.

Azure build-up service · 5.0 · 2025

I highly recommend Banu as a Power BI and Power Query analyst. Good technical skills, professionalism, and a strong commitment to delivering outstanding results.

HR data analytics and process · 2023–24

Attentive to details, knowledgeable, able to deliver on time, very responsive.

Power BI and Python financial dashboard · 5.0 · 2021

Skills

Grouped by where they sit in the platform.

Microsoft Fabric

  • Lakehouse, Warehouse, Delta tables
  • Data Factory pipelines, metadata-driven
  • Spark notebooks (PySpark)
  • Semantic models: DirectQuery, Direct Lake
  • EventStream, EventHouse, KQL
  • Translytical Task Flows, Data Agents
  • Capacity and cost automation

Azure and platform

  • Data Factory, Synapse, Databricks
  • Azure SQL, ADLS, Functions
  • Azure DevOps CI/CD, GitHub Actions
  • Landing-zone and IaC patterns
  • Entra ID and access management
  • PostgreSQL, REST APIs, Delta Sharing

Analytics and training

  • Power BI, DAX, T-SQL
  • Star schema and semantic model design
  • Python, Power Platform
  • Hands-on Fabric and Power BI training
  • Trainer guides for mixed-skill teams

Governance and security

  • Microsoft Purview in platform architecture
  • Lineage and classification at scale
  • Sensitivity labels and DLP
  • Information protection (SC-401)
  • Workspace security, data-quality frameworks

Certifications

All current and verifiable on LinkedIn.

  • DP-700Microsoft Certified: Fabric Data Engineer AssociateJuly 2026
  • SC-401Microsoft Certified: Information Security Administrator AssociateAugust 2025
  • AZ-104Microsoft Certified: Azure Administrator AssociateJanuary 2025
  • AZ-900Microsoft Certified: Azure FundamentalsNovember 2023

About

I started as a data analyst, building reports and answering questions for business teams. The harder problems kept moving upstream: data that was late, duplicated or impossible to trust. So I moved upstream with it and became the person who builds the platform.

Today my sweet spot is end-to-end Microsoft Fabric: ingestion pipelines, medallion lakehouses, star-schema warehouses, semantic models and Power BI delivery, with Azure DevOps for CI/CD and Purview for governance. I also train the teams who inherit what I build, from business users writing their first measure to IT leads adopting Fabric practices.

Independently, I'm a Top Rated Plus freelancer on Upwork with 50+ projects, a 100% job success score and over $200K earned, mostly as a long-term partner rather than a one-off hire. Master's in Management (PSG Institute of Management), BSc Computer Science.

Based in Bengaluru. Open to full-time Fabric and data-platform roles, and to selected consulting and training engagements.

Contact

The fastest way to reach me is email. I reply within a working day.