Answers fromyour own data.

The record so far, in four numbers:

  • 14 years of experience
  • 86 projects delivered
  • 112 models now in use
  • 5 consultants on staff

Sample figures for the Orrery demo.

We turn your own records into forecasts and prices your team can check.

Most clients already have the data: sales exports, a CRM, a stock system and some spreadsheets. We clean it once, model it in SQL and Python, and hand the code to your team.

“Every number we send comes with the query that produced it.”

Our services

Six kinds of work. Each one starts with a written scope and a fixed price.

Two of our last five projects ended with a spreadsheet, not a model. If a pivot table answers your question, we say so in week one and bill for week one. The rest of the budget stays with you, and so does the spreadsheet. We would rather lose a project than sell you a model you do not need, and we put that promise in writing.

Book a call

How a project runs

  1. Diagnosis first

    Two weeks with your data and the people who use the numbers every week. You get a ten-page memo: what the data can answer, what it cannot, and three options, each with a price.

  2. A single owner

    Each project answers one business question, agreed in writing, with one named person on your side who decides. Any change of scope goes through that person and gets a new price before work goes on.

  3. Your own tools

    Models run in your own warehouse, or as a plain Python script on a server you already own. There is no platform to license, no fee per seat, and no login that stops working on the day our contract with you comes to an end, or after.

  4. Handover & review

    In the last week your analyst reruns everything while we watch and answer questions. Ninety days later we come back for a full day, at no charge, to check that the numbers still hold up in your hands and in your tools.

Have a decision coming up?

Tell us the decision you are stuck on. We will say on one call if data can help.

Get started