Case Study 01
Building a dbt Data Foundation from Scratch: A Case Study for Startups Without a Data Team
Health tech · ~50 employees · No data person, just a CEO, AI, and a lot of manual checks. Now 12 people rely on the same infrastructure.
Results at a Glance
0 → 1
dbt data foundation built from scratch
10 hrs/week
given back to the CEO by eliminating manual data validation
Automated
data quality checks running continuously
The Challenge: No Data Infrastructure, No SQL Expertise
The client was growing fast, but had no data person, no data model, and no reliable way to answer basic business questions. When the CEO needed metrics for internal reviews or external reporting, he'd use AI to generate SQL queries against their PostgreSQL database, then manually cross-check the results because he couldn't fully trust what came back. It was slow, error-prone, and not something that could scale. There were no dbt models, no documentation, no automated quality checks. Just raw operational tables and a founder doing his best to extract truth from them.
How We Built a dbt Data Foundation from Scratch
We started by understanding the business from the ground up. Who their clients were, what they delivered as a product, what numbers actually mattered. From that, we designed simple, business-friendly dbt models that mapped directly to how the company thinks and operates, not just how the data happened to be stored in PostgreSQL.
Once we had the right data model on paper, we worked backwards into the database to make it a reality, fitting the existing data into the right shape with the necessary transformations.
Along the way, we uncovered a range of data quality issues that had been quietly skewing their metrics. We documented every one of them, traced each to its root cause, and resolved them together with the development team.
The final dbt project included:
- Mart-layer dbt models built around real business concepts: clients, products, and key business events
- Calculated columns for the KPIs the CEO was tracking day-to-day
- dbt tests for automated data quality monitoring, so problems surface on their own rather than after a manual spot-check
- Full dbt documentation so anyone joining the team understands what every model means and where the data comes from
A deliberate decision was made early on to keep the stack lean and manageable without a dedicated data team. The client runs on the free tier of dbt Cloud, which covers everything they need and requires no internal maintenance or setup overhead. This foundation also became the backbone for the self-serve analytics layer built on top of it with Metabase. Read that case study here.
Results
- The CEO still uses AI to query the data, but with solid dbt documentation and a clean data model underneath, the results are reliable enough that manual cross-checking is a thing of the past
- What used to be a one-person workaround is now shared infrastructure. 12 people across the company rely on the same dbt models to do their work
- Data quality issues that were silently distorting reported metrics have been identified and resolved
- dbt models now reflect how the business actually operates, not just the shape of the database schema
- Automated dbt tests catch data issues before they ever reach reporting
- A fully documented dbt project, ready for a first data hire to take over and extend
“They structured our dbt, reduced platform costs, and left documentation so thorough our team kept building on it. No dependencies, no technical debts. More than a one-time delivery, it became the foundation for reliable metrics and data-driven decisions we're still evolving today. Professional, collaborative, and genuinely focused on long-term value.”
What's this costing your company?
Run our 2-minute calculator and find out where your company stands.
Working with a similar challenge?
Book a free 1-hour audit call and we'll tell you exactly what we'd build and why.
Book Your Free Audit Call →