What we doThe Mansions · Data engineering consulting

Data architecture.

We find every place your data lives, from spreadsheets to inboxes, then put it all in one place, organised and connected, so your data is reliable.

Not sure where your data stands? The quiz shows you in five minutes.

What data architecture and data engineering consulting actually are

Data architecture is the plan for where every piece of your business's information lives and how it all connects. Data engineering is the work of building it, so your client records, files and numbers stop living all over the place and end up in one. That's the work we do, and it comes in three parts.

  1. The plan

    The first is the plan, the data architecture itself. It decides which copy of each record is the one everyone uses, and how each client connects to their documents and invoices.

  2. The pipes

    Then there are the pipes, the data engineering part. They carry your data from the tools you already use into that one place and back out again, on their own, so nobody has to type the same thing twice.

  3. The readout

    And then there's the readout, which is where data analytics comes in. It's the reports and answers you get from that one place, and you can see where each figure came from.

AI can only work with the data you give it. Once your data lives in one place and it's reliable, every automation, agent or AI you build on top of it starts from the real picture of your business.

Every business keeps its data a little differently, but the work runs through the same five steps, from finding it all to building on top of it. At each one, part of the work is ours, and part of it should always stay with your people.

Step 01 · Data audits

Most businesses have a lot more data than they realize, spread across a shared drive, a few spreadsheets, the inbox and, quite often, somebody's head. That's why we think all data consulting should start with a data audit.

We read every source view-only, which means nothing changes while we look. We count what's there, find what's doubled, missing or saved in different ways, and check where two sources tell a different story about the same thing. AI does the heavy reading, and a person checks what it finds.

You get a map of every source, what it holds and where the copies clash, with the fixes in the order that recovers value soonest.

And if the real problem turns out to be how work moves between people, that's where our operations consulting comes in.

Stays with your team

The answers only your people know, like which file is the real one, what a field means and whether two clients are really one.

Step 02 · Cleanup and migration

The same client gets typed a few different ways, the same invoice is saved twice, and there's a total that nobody can quite prove.

We take out the duplicates, and every merge can be undone. Where a value is missing, we recover it from your own documents, filling blanks only and never making one up. Anything unclear comes back to you as a plain question.

Then we move it all to its new home, with the old source left read-only and the counts checked before and after. That's what people mean by data cleansing and data migration.

Stays with your team

The call on every merge that isn't obvious, and what gets kept, archived or let go.

Step 03 · One source of truth

Each of your tools probably keeps its own copy of your clients, so the same person can end up with a different phone number in each one.

So we give each kind of information one home, which is all a single source of truth really is. Each client, supplier, document and invoice lives in one place, and everything else points to it.

Then comes data integration, where we connect your tools to that home so they can read from it and write back to it. We also set up who gets to see what.

Stays with your team

Who sees what, and what your own words mean, like when a client counts as active or an invoice counts as paid.

Step 04 · Business intelligence

Every tool you use comes with its own dashboard, and each one is judging its own work.

So we build your reports on the one source instead, where every figure traces back to where it came from. Anything the system works out for itself is marked as worked out, and never blended in with a real figure. The reports refresh on their own, and AI can draft the summary.

You'll often see this called data analytics consulting. For a small business, that work usually starts with the handful of numbers you already check every week.

And if you want a dashboard your own clients log into, that's our custom software work.

Stays with your team

What the numbers mean, and what you do about them.

Step 05 · Systems built on top

Once your data has one home, everything else can be built on top of it.

Automations can move the work along on their own, and agents can read the record before they act. You can even ask an AI about any part of your business and get an answer that comes from your real numbers.

We build in that order on purpose, the data first and then the systems, whether that's automation or AI built on your own data.

Stays with your team

Which system gets built first, and the sign-off on what it does.

Data work we have done

Every one of these is real work, for a real business. The names are left out, and the numbers come from our own records.

Built

A €3.37M pipeline that was really €2.76M

An owner-run contractor

Its quotes and client records sat in 3,030 scattered Drive items, and together they added up to a €3.37M sales pipeline. We read every one view-only, rebuilt the clients, jobs and quotes in one place and took out the duplicates, checking every unclear one with the owner. After the rebuild the real number was €2.76M, and €612K of the old total was duplicates. Along the way, 316 of its 328 quotes ended up carrying a value, recovered from the company's own documents.

Built

Four tools' numbers, gathered in one place

An email marketing agency running a portfolio of DTC brands

Its clients' numbers lived in four places, from the email platform and the commerce platform to web analytics and product analytics, each with its own export. We built one dashboard to gather those four data sources into one shared structure with a daily sync, on the agency's own accounts and under its own name. We wired the first brand into it and then handed the whole build over.

four data sources

Read the dashboard case

Found

Ad reports that didn't add up

An owner-run e-commerce brand

We took the brand's commerce, ad, web analytics and email data apart, one source at a time, then made them answer to each other. Its ad platform was claiming 1.78 times the sales its own web analytics credited to the ads on last click. In the same data, we found €423K in potential revenue sitting in a list of past customers it already owned. When a separate pass re-checked the email findings, all 38 held.

Built

A checked list first, then the agents

An insurance brokerage

We turned two public business registers into one list, with the details the registers' own download leaves out, and checked it against the registers' own counts. Then we built two agents on top of it, one to pick up new businesses every day and one to refresh the rest on a schedule.

How we work

It always starts with our free consultation. We study how your business really runs and which tools your data lives in, then hand you the plan, with every problem priced and every fix designed. From there, the decision on what to build is yours, and once something's live, we stay on it and keep improving it.

See how the free consultation works

What you own, and what drives the cost

You own all of it

  • Your data, tools and builds. What we build is yours, so nothing stops working if we stop working together.

  • The report. Before anything is built, you hold the report, with every problem priced and the fixes put in the order to build them.

What drives the cost

We don't publish prices, because every business is wired differently. Four things move the cost of a fix:

  • The sources. How many places your data lives, and whether each tool lets us in.

  • The cleanup. How much is doubled, missing or saved in different ways before we start.

  • How much there is. How many records you have, and how many years of them need to come across.

  • What it's for. What you want from it once it's in one place, like reports, live dashboards or AI working from it.

Then there's what it costs to run, like software subscriptions and the AI usage a fix needs each month. The report prices every fix, to build and to run, next to what it gives back in hours or money, with the working shown.

Two ways to work with us

  1. Project by project.

    You pick what to sort first from the report, we sort it, and you pay for that project.

  2. An AI subscription.

    Monthly access to the whole team, bringing your sources in one at a time and keeping your data clean as the business grows.

Is data architecture what you need?

Not every problem with your data needs a data architecture project. Find the line that sounds most like you, and you'll see what you actually need.

If this sounds like youWhat you need
Your numbers live in a drive, a few spreadsheets, several tools and people's heads, and nobody's quite sure which report to believe.
One source of truth, built around how you already work. That's what we do.
Your data already sits in one system, and all you want is a dashboard on top of it.
Someone to set up a reporting tool on that system. It's a much smaller job, and worth knowing before you spend anything.
You need help with servers, backups, email or devices.
An IT firm, who will serve you better than we will. We only touch IT where the AI needs it.
You're looking for a data engineering job, or a contractor to sit inside your team.
A job board or a staffing agency, because we don't place people. We build the systems ourselves.

And if you book a free consultation and we're not the right fit, we'll tell you on the call, and you'll still leave with pointers to get organised and move forward.

Frequently asked questions

A data engineering consultant builds the pipes that carry your data from the tools you use into one place. Those pipes keep it all in sync, so your reports and systems can run on it. Ours starts with a view-only audit of everything you already have, then works through the other steps on this page, up to the systems built on top. A person checks the work at every step.

A single source of truth means each piece of information your business runs on has one home, and every person and every tool reads from it. Each client's record lives once, and every invoice and document about that client points back to it. For the email marketing agency on this page, that meant gathering four data sources into one dashboard, on the agency's own accounts.

Business intelligence consulting means turning the data you already have into reports you can make decisions on. It usually covers choosing the few numbers that matter, connecting the tools they come from and building the reports on one source. Ours starts by checking those tools against each other, because every tool reports on its own work. For the e-commerce brand on this page, that check found its ad platform claiming 1.78 times the sales its own web analytics credited to the ads on last click.

A data strategist writes the plan for what data a business needs and how it'll use it, and the work often ends in a document your team then carries out. If a written plan is all you're after, a strategist is a good fit. Our data work starts with a map of your data too, then carries on into the build, one fix at a time.

No. Your data needs to be good enough for the decisions you make with it, and that's rarely the same as perfect. We clean up what your first build needs, then the next area when the next build needs it. If you want to see where your data stands today, the data part of our AI readiness guide shows what to look for.

Not always. We'd rather join up the tools you already own than sell you more of them. Sometimes a tool you already pay for can be the one place everything lives, and sometimes that place needs a build. The report tells you which before you spend anything, and either way it lives in your own accounts.

Start with one area, like your invoices or your client files. Give an AI assistant access to that folder, ask it to go through everything and propose a structure, and check what it suggests before you let it move anything. Once that area is tidy, the next job is connecting it to the tools you use every day, and that's the part we build.

CDMSIN
  1. 01Data audits
  2. 02Cleanup and migration
  3. 03One source of truth
  4. 04Business intelligence
  5. 05Systems built on top

So what would you decide differently with numbers you can trust?