One engineer, working directly on your Power BI

pbiexpert is my independent Power BI engineering practice. I work on the semantic models, DAX, development workflow and Fabric capacity behind your reports. There is no agency and no hand-off: the person you talk to is the person who does the work.

Piyush Jain

Piyush Jain

Power BI & Fabric Engineer · 8+ years in Power BI

I'm an analytics engineer with 8+ years in Power BI, working remotely from India with teams in the US and Europe. Most of my recent work is large enterprise semantic models: hundreds of tables and measures, path-based and table-driven row-level security, role-aware reports and HTML/SVG visuals built in DAX. I set up pbiexpert to do that work directly for the teams that need it.

Worked with: Healthcare & pharma · Consumer retail · Marketing & sales analytics

Results

  • Merged three separate reports into one access-controlled report for reps and managers
  • Embedded Power Apps and Power Automate in reports, saving about 30 minutes of manual work a day
  • Delivered 75+ dashboards used by 500+ business users

What I do

Semantic models

Star schemas, relationships, date tables, row-level security and model size — the foundation every report depends on.

DAX

Measures that return the right numbers and stay fast: filter context, iterators, time intelligence and query-plan tuning.

PBIP / TMDL and Git

Moving teams from binary .pbix files to text-based projects with source control, pull-request reviews and automated checks.

Microsoft Fabric

Capacity sizing, Premium-to-Fabric migration, Direct Lake design and Copilot readiness.

Workflows

Power Apps and Power Automate around your reports: data entry, approvals, alerts and refresh orchestration.

How I work

01 // PRINCIPLE

Models over visuals

A healthy semantic model makes reporting fast and self-service. I fix the data foundations before touching chart colours.

02 // PRINCIPLE

Models are software

I treat models as production code: text-based TMDL definitions, Git, pull-request reviews and repeatable releases.

03 // PRINCIPLE

Performance by design

I keep column cardinality low, relationships simple and DAX in the storage engine, so models stay fast as they grow.

04 // PRINCIPLE

Evidence first

Findings come from your model's metadata and measurements, not opinions. Every recommendation says why and what it changes.

How to work with me

  1. STEP 1

    Start with a free tool

    Check a measure, a model or your Fabric capacity in the browser. Nothing is uploaded.

    Browse the tools
  2. STEP 2

    Get a Health Check

    A fixed-scope, paid review of your model metadata with a written report and a prioritized remediation plan.

    See the Health Check
  3. STEP 3

    Fix it

    Implement the plan yourself, or have me do it in a fixed-scope sprint or an ongoing retainer.

    See how I work

Tell me what is slow or fragile

A short call is enough to decide whether a Health Check, a sprint or just a free tool is the right next step.