Skip to main content

Meet us at HLTH 2026 · Las Vegas · Nov 15–18

Book a 30-minute slot →
← All case studies
Customer engagement
Data warehouse BI AWS

Improving business decision-making through data-warehouse-powered BI

Client
Personalized marketing platform
Segment
Customer engagement, martech
Scope
Data warehouse, ETL pipelines, BI layer
Stack
AWS Redshift, AWS Glue, BI dashboards
Data
High-volume, multi-source, partly unstructured
Refresh
Automated, regular updates
Colleague pointing at data dashboards on a curved monitor while an analyst reviews them
01 · The context

Data volumes that keep doubling

The amount of data generated worldwide by all kinds of systems has been explosively increasing in recent years. And even though this growth has already been astounding, the projections show no signs of stopping – the volume of data available worldwide is expected to double in the incoming years.

But the data alone is not that meaningful. It requires proper processing/analytics/reporting in order to truly harness its power and derive valuable business insights from it. Multiple enterprises and organisations are struggling with that aspect, especially due to specific technical challenges when dealing with high volumes and unstructured data.

02 · The challenge

Significant loads of data from many sources – and no unified reporting

datarabbit regularly advises on a regular basis on such topics. As a part of one of our engagements, we were designing and implementing a system exactly addressing such problems and needs. Our partner, as a platform providing personalised marketing solutions to increase the customer engagement of their clients deals with significant loads of data from a variety of sources. But in order to bring their strategic insights to the next level, both for themselves, as well as their customers, they needed a system, which would process and aggregate the data as well as provide unified reporting and analytics – and which datarabbit prepared.

03 · What we built

Two parts: a warehouse with ETL, and a BI layer on top of it

Our team did that, by building a solution that consisted of two parts. The first one was responsible for data processing and aggregation. We created a data warehouse, with storage and computing capabilities capable of efficiently processing data coming from all sources of interest. Later, these sources were connected with the warehouse through ETL pipelines, which ensured that the data is automatically updated on a regular basis and always available to the final consumers in the recent enough version. These were implemented utilizing AWS Redshift and Glue technologies. The second part was a BI layer, that was connected to the already established warehouse and using it as the underlying backend (together with data present in it) provided analytics, reporting, and business insights through a series of informative dashboards and charts we configured.

  • Processing & aggregation – Data warehouse on AWS Redshift; ETL pipelines on AWS Glue keep every source refreshed automatically.
  • BI layer – Analytics, reporting and business insights through configured dashboards and charts – backed by the warehouse.

BI dashboard in Apache Superset: daily user views and clicks with totals and averages, broken down by gender, age and marital status

Planning a data or AI project in healthcare?

Book an intro call: an hour with an architect about where you are, whether we are the right fit, and what a Discovery Phase would cover. Honest, even if the answer is no.

Nov 15–18 · Las Vegas

HLTH 2026

Booth 127-24, Exhibit Hall. Book 30 minutes before the show: challenges and best practices for AI and data projects.

Book a slot at HLTH