dbt and data warehouse consulting.
Reporting is only as reliable as the models underneath it. When every analyst writes their own version of revenue or stock, the numbers drift apart and nobody can say which one is right.
We build analytics warehouses in dbt on PostgreSQL, BigQuery, Databricks, or Snowflake: tested, documented models with one definition per metric and a trail back to the source systems. Then we write the SQL for the numbers a business actually runs on, from procurement and goods received to stock, shelf life, and reporting timeliness.
What our data warehouse work covers
- dbt projects: staging, models, tests, and documentation
- Warehouses on PostgreSQL, BigQuery, Databricks, or Snowflake
- Pipelines from operational systems, extracts, and spreadsheets
- Operational SQL reporting: procurement, goods received, inventory movement, shelf life
- Data quality checks: completeness, timeliness, and late-arriving extracts
- Single, agreed metric definitions shared by every report
- Connecting systems that were never built to talk to each other
When companies call us in
Every report has its own logic
Revenue in one dashboard does not match revenue in another. We move the logic into tested models so every report reads the same definition.
Data arrives late or incomplete
Extracts land on different days with missing fields. We build checks that say which extracts were late and which fields failed, before anyone reads the numbers.
Outgrowing spreadsheets
The workbook has become the system. We move it into a warehouse the company owns, without losing what the workbook got right.
How the engagement runs
- 01Diagnose. We name the decision, and say what is already good enough to act on.
- 02Analyse or build. Analysis, software, or both. We write down time and scope before starting.
- 03Ship. A read people can use, or a system the company can run.
- 04Review. We sit with the people who decide, then agree what to watch next.
See the full process, an example of the work, or who does it.
Common questions
Which warehouse should we use?
Usually whichever one the company already has, or can run most cheaply. We work across PostgreSQL, BigQuery, Databricks, and Snowflake, and we advise on the choice rather than pushing one platform.
Do you only do the engineering, or the analysis too?
Both. The same practice that builds the warehouse does the analysis it exists for, so the models are shaped by the decisions they have to support.
How much does it cost?
We quote after a short diagnostic conversation, not from a price list, because the right shape of the work depends on what is actually broken. Time, scope, and stop conditions are written down before the work starts. There is no quote, and no charge, for the first conversation.
Who owns what you build?
The company does. Code, models, dashboards, and documentation are handed over. It is not a product you rent from us or a service you have to keep paying us to run.
Related services
If your numbers do not add up, a report is still built by hand, or your team has outgrown its tools, get in touch.
A few sentences on the situation are enough. We reply from the same address you see here.