How do you migrate a data feed safeguarding £34bn of pensions without disruption or downtime?

Is there a more nerve-wracking prospect for a Chief Information Officer than a complex, business-critical data migration project? And when downtime is intolerable and 100% data consistency is a mandatory, this task takes on even more foreboding dimensions.

For Railpen, who manage £34billion in pension assets for more than 350,000 rail workers, replacing a data feed powering more than 70 critical applications within a fixed cutover window was a task that had to be handled with care, precision and, above all, an understanding of the risk of error. Any failure to migrate would be multiplicative across all 70 systems simultaneously, with potentially catastrophic consequences for the financial security of Railpen and its members.

Before any build began, Softwire conducted a thorough analysis of Railpen’s data estate — mapping every data flow, every integration point, and every downstream dependency across the 70+ connected applications. This analytical foundation was the mechanism by which the team identified the migration path that would minimise downstream disruption and de-risk the cutover.

Working as a single team, Softwire, the supplier of Railpen’s new data platform and Railpen themselves ran the new platform in parallel with the existing feed, implementing Azure Functions to ensure that the new platform met future needs and didn’t simply mirror existing functionality.

This was not a lift-and-shift of the stored procedure logic which had been embedded across thousands of lines of difficult to read, hard to change and invisible to governance SQL. Instead the new platform and the ingestion layer that underpins it are visible, versioned and fully auditable – a data layer with end-to-end lineage built in.

While we maintained a rollback plan throughout the migration process, through the efforts of the combined team this was not actioned at any point.

And beyond the delivery, this programme produced a material step forward in the quality and resilience of Railpen’s data infrastructure, improving observability and traceability and enhancing Railpen’s data management capabilities.