Solution

Quantiphi successfully migrated and re-engineered 5 key dashboards to Looker Enterprise on Google Cloud. These dashboards were developed using LookML and powered by dynamic data from the BigQuery instance.
To enhance the utility of these dashboards, Quantiphi implemented an on-the-fly time series forecasting model that enables business users to generate tailored forecasts directly within the Looker interface. This system integrates templated filters in Looker with dynamic training data passed through BigQuery and into a custom forecasting model deployed on Vertex AI.
As part of the dashboard modernisation effort, forecasting capabilities were initially implemented using BigQuery ML’s ARIMA_PLUS model to support early-stage decision-making. The approach was later scaled with a more flexible, customized solution using the Facebook Prophet model.

Results

Developed and Deployed Advanced Looker Dashboards: Successfully built and launched five dynamic Looker dashboards featuring 29 critical KPIs, delivering effective patient flow and service demand analysis and forecasting for Aneurin Bevan University Health Board.
Integrated Dynamic Forecasting Capabilities: Implemented a robust and dynamic forecasting solution using Facebook Prophet hosted on Vertex AI, integrated directly into Looker dashboards with execution time of 10 – 20 seconds. This allows users to generate tailored forecasts on-the-fly based on selected filters, moving beyond static models for more accurate and actionable insights into patient flow.
Developed a Scalable, Centralized Analytics Platform: Implemented Looker on Google Cloud Platform (within NDR) as a modern, centralised reporting and analytics solution for ABUHB, designed with the scalability to expand across all of Wales in the future.

Testimonial

“For me, it was an excellent project to work on. Quantiphi delivery team was steady and calm, and colleagues really appreciated onsite attendance at our office, and ability to talk through questions and issues, which I think was a core success factor.
They were always available and responsive to any questions or issues raised, and again were very calm and accommodating to make sure project stakeholders were heard and requirements were satisfied.
The wider team were very open to feedback and actively reached out to seek input from ABUHB and myself, and found great solutions to resolve any difficulties like extending support to reflect our planning team’s workload to cover UAT testing/feedback/re-testing.
I think we have a better mutual understanding now of ways of working, and some of the challenges we face in DHCW/NHS Wales in moving quickly on these types of projects. This I think sets us up well for future work.”
Louise Smith, Project Manager, DHCW, NHS Wales