Building a Scalable Data Analytics Platform for Intelligent Health

The Client

Intelligent Health is a UK-based company dedicated to improving public health and resilience by engaging communities in physical activity. Their award-winning program Beat the Street transforms towns and cities into game boards, encouraging residents to move, connect, and stay active. Data analytics is central to measuring impact and driving community insights. 

The Challenge: Unifying Fragmented Data Streams into a Robust Analytics Platform

Intelligent Health faced the challenge of combining and analyzing diverse data streams collected through its Beat the Street program. The initiative generates information from multiple sources, including RFID swipes at Beat Boxes, GPS-triggered mobile app interactions, and in-game surveys, yet the existing approach risked becoming overly complex, costly, and time-consuming to manage. As a mission-driven organization, Intelligent Health needed a solution that delivered robust analytics while staying lean, affordable, and easy to maintain.

At the same time, growing demand for timely insights in the healthcare and community sector added pressure to adapt. Programs like Beat the Street depend on real-time visibility into engagement and impact, but legacy approaches to analytics often demand large-scale infrastructure, specialized expertise, and lengthy rollouts. Intelligent Health needed a scalable, data-driven platform that could provide insightful dashboards quickly, without unnecessary complexity or prohibitive expense, to continue advancing community health outcomes.

The Solution: A Scalable, Lean, and Serverless Architecture to Match Business Needs

The project began with an ambitious architecture proposal based on Cloud Composer, but this quickly proved more sophisticated than the data volume required. Recognizing the importance of simplicity, the team shifted to a leaner, serverless design. This new approach used GitLab CI/CD pipelines for orchestration, Terraform to automate infrastructure provisioning, and a focused set of GCP services—BigQuery, Google Cloud Storage, and Data Transfer—for data processing and storage. To manage transformations, dbt was introduced, with the dbt-external-tables package enabling raw data in GCS to be easily accessed and transformed in BigQuery.

The streamlined architecture struck the right balance between functionality and efficiency. It enabled fast iterations and straightforward troubleshooting, while keeping operational costs low. Orchestration costs were effectively zero, as Intelligent Health typically remained within the free usage tier. The setup was version-controlled, ensuring transparency and repeatability across development and production environments. Later, when Intelligent Health migrated its DynamoDB backend to a single-table schema, the well-structured dbt models allowed the analytics pipeline to absorb the change with only minor updates, leaving dashboards and reporting unaffected. This adaptability highlighted the strength of the chosen methodology: start small, stay flexible, and provide a scalable solution that the client could continue to expand independently.

The Results: Rapid Delivery, Smooth Migrations, and Cost-Efficient Operations

The collaboration quickly produced results that gave Intelligent Health both speed and long-term value:

What's Next

The collaboration continues on demand with minor enhancements and new dashboards. The solution is robust yet flexible, positioning Intelligent Health to scale data maturity across future community programs.

Interested in learning how we can help you build a modern, scalable data platform in just weeks? Get in touch today.