Fractional Data Architect on Thomas Nys | Fractional Data Architect

Companies are writing one job ad for three different jobs and calling it an AI data engineer.

A broken pipeline costs you an hour.

NIS2 turned one client’s logging gap into a board-level problem inside a week.

A dataset with no owner and no cost ceiling will always be somebody else’s problem.

The hybrid teams paying the least tax have two rituals and fewer meetings.

Senior data engineers quit stacks that page them at 3am, and the exit interview says “growth”.

The short version A useful data platform assessment answers 3 questions: What is solid enough to keep?

The short version Data architecture due diligence should establish whether the platform can support the business plan without hiding material security, reliability, cost, ownership, or delivery risk.

Book a discovery call Use the 30-minute discovery call to explain what is happening in your data platform or team.

Nobody wants to be the one who questions the tool the last person bought before they left.

Every platform vendor offers 20-30% for an annual commit.

A backend deploy broke the data pipeline for the third time this quarter.

A client built the same data inventory twice in 18 months.

Postgres handled it fine.

We changed the questions in the postmortem, and the recurring outage stopped recurring.

Data platform changes have a good window and a bad one.

We scored a client’s pipelines on cost against consumption.

Four questions, asked consistently, replaced most of one engineering manager’s architecture decisions.

Duplicate invoices came from the recovery, not the outage.

A company at 40 people needs senior data judgment about 2 days a week.

Single-cloud is a fine strategy.

Nobody owned the orders table.

Your engineers are already learning AI at work.

A failure that cost one dashboard last year now takes out 40 downstream jobs.

Half the “Head of Data” searches I see are two different jobs wearing one title.

Every AI roadmap I review gets the same first check: whether the data could go in a board deck unchecked.

One pipeline quietly consumed a full data engineer.

A 60-person scaleup asked me to design their data mesh.

An ETL cost audit cut a small team’s bill by roughly a third.

Tech debt gets a backlog and a budget.

DORA has applied since January 2025.

On one team, engineers reported their own near-misses weekly.

BigQuery vs Snowflake for ad-hoc work comes down to one choice: pay per scan or pay per minute.

The DORA metrics are over a decade old, proven on software teams, and unused on most data teams I meet.

Before the term sheet, a CTO asked for a modularity audit.

The data lake had 300 datasets and no owner list.

A client budgeted 2024 salaries for a 2026 data engineer role.

An auditor asked who could read the customer table.

A CTO said “that’s exactly us” four minutes into the call.

A team of 8 data engineers spent roughly 30% of their week negotiating schema changes in Slack.

Every failed RAG project I’ve reviewed had the same upstream issue: nobody modeled the data underneath it.

The warehouse that carried you to Series A can quietly become the wrong one by Series B.

We changed one screening step and two senior data engineers are still there, past year two.

We mapped NIS2, DORA and GDPR incident duties for one client.

Your newest data consumers don’t file tickets.

A “temporary” sync script ran for 18 months.

The Phoenix Project is 13 years old.

By 20 engineers, most data teams have one person every change waits on.

You can measure architecture health by the price of an exit.

A scaleup added 3 data engineers to catch up.