Reporting off production
Dashboards query the live database replica, slow it down, and break whenever an engineer renames a column.
Home / Packages / Data warehouse
Package 04 · Fixed price
Data warehouse
Your sources landed in BigQuery or Redshift, modelled with dbt into tables that match how the business talks, tested every run, documented, and scheduled daily with alerts. One place every report starts from, in your own cloud account.
from $3,500Fixed price, agreed up front
2 weeksDelivery
Your cloudRuns in your account
Who it is for
You will recognise one of these
Scripts nobody dares touch
A folder of cron jobs and notebooks that one person understands. When they are on leave, reporting stops.
One definition per metric
Finance, sales and ops each compute 'active customer' differently. You need one agreed number.
WHAT YOU GET
Everything in the package
Delivered in 2 weeks, in your own cloud account, with the code and documentation handed over.
- A warehouse on your Google Cloud (BigQuery) or AWS (Redshift) with raw, staging and mart zones and a naming standard
- Ingestion for up to five sources: your application database, ad and marketplace APIs, spreadsheets, SaaS tools
- A dbt project: staging models, marts for your main reporting needs, and tests for uniqueness, nulls, accepted values and freshness on every run
- Generated documentation so anyone can find what a table means and where it came from
- Daily scheduled runs with alerts to Slack or email when something fails or data is late
- Cost controls: partitioning, clustering and a budget alert, with roles so people see only what they should
- Everything in your Git repository with CI, a two-hour handover and recorded training
How it works
4 steps, 2 weeks
STEP 01
Design
Days 1 to 2. Sources, the metrics that matter, and the zone and naming standard, agreed in writing.
STEP 02
Land the data
Days 3 to 5. Ingestion for each source into the raw zone, then staging models that clean and type it.
STEP 03
Model and test
Days 6 to 9. The marts your reports need, tests on every model, docs generated, dashboards pointed at the marts.
STEP 04
Handover
Day 10. Training, the runbook, and your team running a full pipeline on their own.
Scope and the small print
- The price covers up to five sources and roughly fifteen mart tables. Larger scopes are quoted before we start.
- It runs in your cloud account. Warehouse running costs at startup scale are usually between $50 and $300 a month.
- Works with whatever BI tool you have: Metabase, Looker, Tableau or Power BI.
- Change data capture and streaming are quoted separately; this package is daily batch, which is what most companies need.
Questions
Asked before buying
BigQuery or Redshift?
Usually whichever cloud you already use. If you are on neither, BigQuery is cheaper to run at small scale and simpler to operate, and we will say so.
Can you migrate our existing scripts and reports?
Yes, within the five-source scope. Existing SQL is usually the fastest way to learn what the business needs.
Do we need a data engineer in-house afterwards?
Someone who can read SQL and run a documented command is enough. The project is built so a developer can extend it, and we offer monthly support if you would rather not.
How is this different from the Metabase package?
Metabase is the window; the warehouse is what it looks at. Small teams start with Metabase on their production replica and add the warehouse when the reports outgrow it.
Let's talk
Not sure it fits? Ask.
A 30-minute call, no deck. We will tell you whether this package is the right one, what we would do first, and whether you even need us.