Technology Enablement

Description

SWL ICB is embarking on the development of a new data platform using Microsoft Fabric with an agile, iterative development approach. While the outcomes are aspirational rather than guaranteed, the project is targeting key objectives: Expedited Data & Insights: The platform aims to ingest live patient identifiers from hospital providers every 15 minutes, generate real-time insights, and update Power BI dashboards on the same interval. In practice, this goal will be refined over successive agile cycles to ensure the system meets the required performance for feeding data back to providers, GPs, and community partners. Enhanced Innovation: The initiative envisions hosting a population health management (PHM) data model accessible to all ICS analysts. This would empower teams to explore insights and build machine learning models independently. The agile methodology ensures that feedback and evolving requirements can drive continual improvements and innovative solutions. Power BI Reporting on Fabric: As part of testing, SWL ICB will migrate its existing Power BI reports onto the Fabric tenancy. The objective is to improve connectivity and reporting performance-ideally increasing the refresh frequency for select reports to every 10 minutes. Given the current throughput of 300 users and 3,000 monthly reports (with around 50GB of PBIXs), initial iterations will focus on stability and scalability, with subsequent refinements based on real-world performance data. Modernized SQL Environment: The plan includes consolidating roughly 3TB of SQL databases within Fabric to remove existing tenancy, server, and supplier constraints. The vision is to create a controlled environment with read/write access for ICS colleagues. Agile delivery will help align compute capabilities to the needs of approximately 20 heavy users while adapting to demand as the system evolves. Improved Information Governance and Automation: Leveraging Microsoft Purview, the project intends to monitor data assets closely to support an expanding user base. Additionally, automation-covering traditional transactional tasks and Python-based machine learning package processing-will be developed iteratively to streamline operations and manage complexity.