01
Data platform
Warehouse design, ingestion, dimensional modelling, and the test suite that keeps the numbers honest once nobody is watching. Usually replaces a chain of scheduled exports nobody wants to own.
8–16 wks
Data, systems & webEst. 2019Remote — United StatesOne engagement slot — Q4 2026
Data systems
your team can
still run after
we leave.
Maes Consulting is an independent technical practice. We come in when the spreadsheets stopped scaling, the nightly job started failing, or nobody can say which number is correct, or the website stopped bringing in work — and we leave behind something your own people maintain.
- Engagements
- Fixed scope, fixed price
- Typical cycle
- 2–16 weeks
- Roster
- Two clients at a time
01
Practice
Seven things, done properly
02
Systems integration
Making the tools you already pay for talk to each other properly — with idempotency, retries, and a reconciliation report, rather than a webhook and hope.
4–10 wks
03
Automation & internal tools
The manual step somebody performs every Tuesday, removed. We scope narrowly, ship to one team first, and only generalise once it has survived a month of real use.
3–8 wks
04
Technical due diligence
An outside read on a codebase, a vendor contract, or a build-versus-buy decision. You get a written assessment with the risks ranked and priced, not a verdict.
1–3 wks
05
Website design & build
A fast, accessible marketing site written in plain language, with an editing setup your team will actually use. No page builder you are locked into and no monthly fee for keeping it online — you own the code, the domain, and the hosting account.
3–8 wks
06
Search & AI visibility
Technical SEO, structured data, and content built so search engines and AI assistants can find you, understand you, and cite you. We fix what is stopping you from being found before we write anything new.
2–6 wks
07
Marketing analytics & automation
Tracking you can trust, a CRM that is kept up to date without anyone typing it in, and email follow-up that runs on its own. The end result is one report that shows which channel brought in the work.
2–6 wks
02
Selected work
3 of 3 shown
001
- Batch duration
- from 9h 12mto 11m
- Manual corrections / mo
- from 340to 6
2025
002
- Source spreadsheets retired
- from 14to 0
- Days to close the month
- from 9to 2
2025
003
- Proposed rebuild budget
- from $1.4Mto $0
- Cost of the work we recommended
- $96K
2026
03
Method
Same four steps, every time
01
Read the system
Before proposing anything we trace the data end to end and interview the people working around the gaps. Most briefs describe a symptom. The first week is spent finding what is actually causing it.
Output — Written findings
02
Name the constraint
One thing sets your throughput. We identify it, size the cost of leaving it alone, and agree in writing what we are and are not solving. Scope disputes happen here, on paper, not in month three.
Output — Scope & fixed price
03
Build the smallest version
Working software in front of real users inside three weeks, then iterated. We do not disappear for a quarter and return with a platform. You see the thing while it is still cheap to change.
Output — Shipped increments
04
Hand it over properly
Runbooks, tests, architecture notes, and two weeks of paired work with whoever inherits it. The engagement is finished when your team ships a change to it without calling us.
Output — Handover & runbook
04
Fit
Read this part before writing
We are a poor choice for roughly half the people who contact us, and it is cheaper for everyone if that surfaces now.
Good fit
- You have a specific system that is failing, and someone internally can say how.
- There is a team — however small — who will own the result afterwards.
- You want the constraint named honestly, including when the answer is inconvenient.
- Budget is approved, or you want help sizing it before you ask for it.
Poor fit
- You need staff augmentation — bodies on a backlog, directed by your managers.
- The goal is a strategy deck. We write findings, but we are hired to build.
- Nobody will inherit the work. We do not take retainers to be a permanent dependency.
- The decision has been made and you want it validated. We will probably disagree.