BLUEHARBOURD3
Book an intro call

Depth 20.2 · Cloud, data and modernisation · Level 3

Data platform

A data platform is the set of tools that collects data from a company's systems into one place, cleans it and shapes it into tables people can query and trust. The central part is usually a data warehouse with a modelling layer: written, tested definitions of what a customer, an order or revenue means, so two reports never disagree because they counted differently.

Specification

Input
Access to the source systems, the reports people use today, and someone who can settle what each business term means.
Method
We load source data unchanged into the warehouse, build models on top of it in version controlled code, test each model for completeness and consistency, and write down every business definition the models use.
Deliverable
Warehouse and modelling layer: ingestion pipelines, tested models in your repository and a glossary of business definitions.
Measured by
Model tests passing on each run, data freshness against the agreed schedule, and reports that reconcile with source systems.

Most often bought inPayments and fintech, B2B software, Logistics and supply chain, Ecommerce

The artefact, before and after

Warehouse and modelling layer: ingestion pipelines, tested models in your repository and a glossary of business definitions.

Before: a blank data platform record with a row for each of ingestion from source systems, data warehouse setup, version controlled data models, data tests and freshness checks, glossary of business definitions, and no entries. After: every row carries a finding and a checked status, which is the state the work hands over.

Inside it

  • 01Ingestion from source systems
  • 02Data warehouse setup
  • 03Version controlled data models
  • 04Data tests and freshness checks
  • 05Glossary of business definitions
Deliverable · Level 4→ Warehouse and modelling layerA warehouse and modelling layer is the place where data from your operational systems lands, is cleaned and is shaped into tables your reports and models read, with each table defined once, tested and documented.

Typical first engagement

The usual order of a first piece of work. Timing and price are set at scoping, once we have seen the stack.

  1. 01Questions firstYou get: The business questions the data must answer, written down with their owners.
  2. 02ModelYou get: A warehouse and modelling layer with documented tables and tests.
  3. 03First reportYou get: One recurring report built on the model and checked against the source.

Cloud, data and modernisation

Next in this line

Ask about this part of the stack

Write one question about data platform in your own systems. We answer it in writing and name what we would need to see to answer it properly.

Send one question

A written reply, no call needed.