Data Nobody Argues With
Strategy, pipelines and reporting on Google Cloud, built so the number in the dashboard is the number in the board pack.
Drowning in data?
Most companies we meet aren’t short of data. They’re short of agreement about it.
Finance has one revenue figure, the product team has another, and both are right according to the query behind them. Someone rebuilt last quarter’s report from a CSV because they didn’t trust the dashboard. The data team has a six week backlog of requests that are mostly people asking the same question in slightly different words, and the board wants conversational analytics yesterday.
None of that is a technology problem, but technology is where it shows up. It comes from having no agreed definition of anything, pipelines nobody owns, and reporting built on top of raw tables.
We fix it in that order:
- Agree the definitions
- Build the pipelines properly
- Then put a governed reporting layer on top
Our Services
What We Do
Data Strategy
Where your data is, what the business can’t currently answer, and the order to build things in. You get a decision document with a build order and a cost estimate. Not sixty slides.
ELT and ETL
Pipelines into BigQuery from your operational databases, SaaS tools and event streams. Tested, monitored, and with a named owner for each one, because the cost of a pipeline is the five years after you build it.
Data Mesh
Domain teams owning their own data products, with a central platform and governance underneath. Worth doing when your central data team has become the bottleneck. A bad idea before that.
Looker
A semantic layer in LookML so revenue means one thing across the business, then governed dashboards and embedded analytics on top of it, with conversational Analytics to keep the executive team happy and take pressure of the data team.
Where to start
If your reporting is disputed and you don’t know why, start with Data Strategy. Two to four weeks and you’ll know what you’re dealing with.
If you already know what to build and just need it built, skip the strategy and start with ELT and ETL. We’d rather build than write about building.
If your problem is that everyone builds their own dashboards and no two agree, that’s Looker, and specifically the modelling layer.
If your central data team is the queue everything waits in, and you have engineers sitting inside the business units, that’s when Data Mesh is worth a conversation.
Migrating a warehouse off another cloud is a different job. See AWS to GCP Migration.
Why us
BigQuery is where we live. We only work in Google Cloud, and we use Databricks where a lakehouse is genuinely the better fit rather than because it’s on a partner list.
We hand over what we build, documented, in your repository. If your team can’t run it without us in six months, we’ve done the job badly.
Take Action now
Let’s create something extraordinary together.
Let's build something transformative together on Google Cloud.
Schedule a complimentary architectural review session with our certified Google Cloud and AI engineering specialists.