BDDC · Ben Dorrington Auckland · NZ
Freelance data analyst · Applied AI

From scattered to trusted

I bring scattered, disconnected data together into one clean, reliable foundation - then build the dashboards, forecasts, and AI automations that turn it into decisions you can move on.

Start a project → See how I work
Python · SQLCore stack
Claude · OpenAIAI toolkit
AucklandWorks globally
Services

What I build

Two core offerings, plus four supporting services that extend the same toolkit into adjacent problems.

Data Foundations for AI

Most AI and analytics projects don't stall on the modelling - they stall on the data underneath: scattered across tools, inconsistent, and unreadable to a machine. I bring your disconnected sources into one clean, reliable layer your team and your AI tools can trust.

  • Consolidating disparate sources - separate databases, analytics, spreadsheets - into one warehouse (BigQuery, Snowflake)
  • Reliable Python & SQL pipelines that keep it fresh, validated, and documented
  • Merged into one clean schema your team, your BI, and your LLMs can all trust

Live Dashboards

Web‑based dashboards your team will actually open. Live, fast, and tied to the questions that matter - not a 12‑tab spreadsheet nobody scrolls past row 30.

  • Custom dashboard apps built for your data - Plotly, or hosted on Databricks
  • Power BI / Tableau where it fits the stack better
  • Built straight on your data foundation - live, refreshed, and easy to extend
Also available

Applied AI & Automation

AI put to work where it moves the needle - LLMs that read and interpret thousands of documents, automated reporting, and custom internal tools your team can use tomorrow.

Forecasting & Modelling

Forecasts, planning models, and market-share & demand analysis - built to explain which way things are heading and why, not just chart it.

Ad Hoc Projects

One‑off deep dives, audits, migrations, or builds. Scoped tightly, delivered fast.

Ongoing Support

Fractional analytics support for teams that need consistent data expertise without a full‑time hire.

About

About BDDC

Ben Dorrington
Ben Dorrington · Auckland, NZ

BDDC is me, Ben Dorrington. A data analyst working solo with the businesses that have data sitting around and no one to make it useful.

Fifteen years inside companies like Philips, FedEx, Z Energy and Chevron - figuring out what data actually changes a decision, and what just sits in a dashboard.

I left full-time work in 2021 because I kept seeing the same pattern: businesses sitting on useful data with no one inside the company who had time to make it useful. BDDC is me filling that gap.

Previously
Philips/Versuni FedEx Z Energy Chevron
Core stack
Python SQL
Also fluent in
Databricks Snowflake Power BI Tableau Excel
AI & Automation
Claude OpenAI
Selected work

Recent projects

Two recent builds - what the data looked like going in, and what became possible once it was brought together.

B2B e-commerce · Web Analytics

One global view from disconnected sites

The same storefront ran across several sites worldwide - each with its own Google Analytics property in a separate cloud project, and its own SQL database. There was no way to see customer behaviour globally. I built pipelines to consolidate every analytics dataset into one BigQuery project, synced the transactional databases into Snowflake, and merged the separate schemas into a single clean model. On that foundation, a Databricks-hosted dashboard app put feature usage, project planning, and other analysis in one place their team could actually use.

BigQuery Snowflake Databricks Python · SQL
Appliance manufacturer · Market intelligence

Understanding emerging and fading products

This brings together three sources that normally live apart: SKU-level market data, product feature specs, and scraped customer reviews. Data science detects which products - theirs and competitors' - are gaining or losing category share, and whether it's driven by promotion, distribution, or demand. LLMs read the reviews to explain what customers actually like and dislike. It all reports out through Power BI as an automated read on which SKUs are emerging or fading, and the reasons behind each.

Data science LLMs Web scraping Power BI
Process

Three steps. No jargon, no decks that gather dust

Step 01

Scoping call (free)

We talk through what you need and what's actually possible. I'll ask direct questions about your data, your tools, and what a good outcome looks like. Free, no obligation. Usually around 30 minutes.

  • No prep needed on your end
  • I'll be straight with you about whether I can help
  • Remote - Teams or Zoom
Step 02

Proposal

Within a week you'll have a written proposal: what I'll do, a clear timeline, and how we'll structure it - fixed-scope for a defined build, or day-rate where the work's more open-ended. Either way, no scope creep and no surprises.

  • Fixed-scope projects or day rate - whatever suits the work
  • Defined deliverables and outcomes you can hold me to
  • Clear on scope, timeline, and cost before we start
Step 03

Delivery

Work gets done on the agreed timeline. You get usable outputs - working code, live dashboards, documented models - not a slide deck summarising what could be built.

  • Regular check-ins so you see progress, not just hear about it
  • Code and configs are yours - nothing locked behind me
  • Documented as we go - nothing locked in my head

Ongoing support available after delivery, pay-as-you-go, no retainer.

Contact

Got a problem worth solving?

Tell me a little about it. I'll come back within a working day with thoughts and next steps.