Numbers your
leadership trusts,
in 90 days.

Flux Data Consulting is Joe Han’s solo practice, based in Everett, WA and working with clients remotely, building data foundations for companies of 50–500 people: pipelines, a warehouse, agreed metric definitions and leadership dashboards, delivered in about 90 days.

I build data foundations for growing companies: pipelines from your core systems, a warehouse your team can query, metric definitions everyone agrees on, and dashboards leadership actually uses.

Fixed scope, fixed timeline, and documentation your team can run with after I leave.

Six places the numbers live today, and the single agreed set the 90-day engagement is built to produce.
SourceHeld as
CRMcustomers
Financerevenue
Productevents
ERPspend
SheetsMonday report
Dashboardsdecisions

After 90 days

One trusted set of numbersAgreed definitions, pipelines from the core systems, and dashboards leadership uses.

Most growing companies hit the same wall.

Leadership can’t get a straight answer to revenue by customer, spend by vendor, or production yield.

  1. 01

    Five systems

    Revenue, customers, spend and production each live in a different tool, and none of them agree.

  2. 02

    A Monday spreadsheet

    Someone stitches the numbers together by hand every week. When they’re out, the report is too.

  3. 03

    Quiet breakage

    A source changes, a number drifts, and nobody notices until it’s in front of the board.

Three phases. About 13 weeks. One fixed scope.

Every project follows the same path, so you know what you get and when before we start.

  1. Weeks 1–2

    Assess

    Find where the numbers come from and agree on what they mean.

    • Inventory of your source systems
    • Map of how data flows between them
    • 5–10 priority metrics with agreed definitions
    • A written plan for the build
  2. Weeks 3–10

    Build

    Put the foundation in place and get the first numbers in front of leadership.

    • Warehouse or lakehouse your team can query
    • Automated pipelines from 3–5 core systems
    • Tested data models and a metric layer
    • 1–2 leadership dashboards
    • Alerts when data is stale or broken
  3. Weeks 11–13

    Hand off

    Make sure your team can run it without me.

    • Documentation and a runbook
    • A walkthrough with your team
    • 30 days of fixes after handoff

Day 1

  • Data in five systems
  • Manual stitching
  • Disputed definitions
  • Silent failures

Day 90

  • One trusted set of numbers
  • Dashboards leadership uses
  • Problems caught early
  • A team that can run it

What I build

  • Automated pipelines
  • A warehouse you can query
  • Agreed metric layer
  • Alerts and tests

I work in the cloud you already run: AWS, GCP or Azure. For new builds I default to Postgres or DuckDB/DuckLake for storage and Airflow, Prefect or Temporal for orchestration, with Presto or Trino at scale.

See the full scope

Companies of 50–500 people with no data team, or one stretched thin.

You have a leader who’s tired of not trusting the numbers, and you need a senior engineer to build the foundation, not a large consulting team.

  1. 01

    Tech and SaaS

    Product, revenue and customer data spread across a growing stack.

  2. 02

    Adtech

    Measurement and privacy-sensitive pipelines, the work I did at Meta.

  3. 03

    Manufacturing around Puget Sound

    Spend, production and vendor data stuck in ERP exports and spreadsheets.

Good signs it’s time

  • You just raised money and the board wants real numbers
  • You're about to hire your first data engineer or analyst
  • You're moving off spreadsheets or a legacy ERP export
  • Leadership argues about whose numbers are right
  • The cloud data bill keeps climbing
Joe Han, founder of Flux Data Consulting

8 years as a data engineer at Meta.

I’m Joe Han, founder of Flux. The patterns I bring are the ones I used on data that had to be right at very large scale, applied at the size you actually need.

About Joe

  1. Meta

    Ads measurement

    Retention, growth and adoption datasets used to run products.

  2. Meta

    Ads privacy and compliance

    Pipelines that had to be right, not just fast.

  3. Meta

    Data center spend

    Purchase-order, invoice and change-order pipelines that give leadership visibility into spend at very large scale.

  4. U.S. Air Force

    Operations research

    Before Meta, I was an operations research analyst, building optimization models and executive dashboards, including a model that placed 1,000+ graduates into jobs each year.

What working together looks like.

Straight answers before we get on a call.

What is Flux Data Consulting?

Flux Data Consulting is my solo data engineering practice, based in Everett, WA, working with clients remotely. I build pipelines, a warehouse, agreed metric definitions and leadership dashboards for companies of 50–500 people, and I deliver that foundation in about 90 days.

Who is a project for?

Companies of 50–500 people with no data team, or one stretched thin. You have a leader who’s tired of not trusting the numbers, and you need a senior engineer to build the foundation, not a large consulting team.

Where is Flux Data Consulting based?

Based in Everett, WA, working with clients remotely. I work with tech and adtech companies, and increasingly with manufacturers around Puget Sound.

Who is Joe Han?

I’m Joe Han, founder of Flux. I spent 8 years as a data engineer at Meta, on ads measurement, ads privacy and compliance, and data center spend. Before that I was an operations research analyst in the U.S. Air Force, including a model that placed 1,000+ graduates into jobs each year.

What is included in the 90 days?

Three phases over about 13 weeks. In weeks 1–2 I find where the numbers come from and agree on what they mean. In weeks 3–10 I put the foundation in place and get the first numbers in front of leadership. In weeks 11–13 I make sure your team can run it without me. That includes an inventory of source systems, pipelines from 3–5 core systems, a warehouse or lakehouse your team can query, 5–10 priority metrics with agreed definitions, 1–2 leadership dashboards, alerts when data is stale or broken, documentation, a runbook, a walkthrough, and 30 days of fixes.

How long does it take?

About 90 days, in three phases over about 13 weeks. The two-week Assess phase can be booked on its own.

What tools do you build with?

I work in the cloud you already run: AWS, GCP or Azure. For new builds I default to Postgres or DuckDB/DuckLake for storage and Airflow, Prefect or Temporal for orchestration, with Presto or Trino at scale.

How is it priced?

One fixed price for the whole project, set in a written proposal after a short discovery call. No hourly billing, and scope only changes if we both agree in writing.

Can we start smaller?

Yes. The two-week Assess phase can be booked on its own. If you go ahead with the build, its fee is credited in full.

Do we have to change our tools?

No. I build on the cloud and systems you already run. Where there’s nothing in place yet, I pick simple, widely used open tools your next hire will already know.

What happens after 90 days?

Your team owns the system, with documentation, a runbook and 30 days of fixes. Some clients keep me on a monthly advisory retainer for new sources, new metrics and reviews.

Who will actually do the work?

I will. I take on at most three projects at a time, so you work with the person who designs and builds it, not a junior team.

When is it not a fit?

A one-off dashboard, extra hands for an existing team, or a project with no executive who owns the outcome. I’ll tell you on the first call.

If that wall sounds familiar, let’s talk.

Tell me which decision is hardest to make today because of data, and where that data lives. The first call is about whether this is a fit.

info@flux-data.com