Case study · August 3, 2026

A Hospital Analytics Platform Built to Survive a System Failure

Toronto hospital evolves from spreadsheets to a governed data platform

The client

A specialty inpatient and outpatient hospital in Toronto, serving roughly 3,000 patients across a dozen inpatient beds and a broader outpatient practice. Before this engagement, the hospital had no consolidated data platform. Reporting ran off manual exports and spreadsheets pulled by hand from six to seven separate clinical and operational systems, with no shared source of truth and no way to answer executive questions without days of manual reconciliation.

The hospital’s Head of Data and Strategy had established the goal: give leaders timely, trusted information for planning, performance measurement, and resource allocation. A central data platform was essential to that strategy. Momo Analytics was brought in to design and build the technical foundation that would put it into operation.

The challenge

A hospital this size cannot staff a data engineering team, so whatever got built had to run under one person’s ongoing attention, not a team’s. Resilience was therefore a core requirement. If a laptop failed, a script broke, or the person who built it left, the hospital needed to know the platform could come back, rather than depend on hope. Disaster recovery had never been a written plan, because there had never been a platform to recover.

The approach

The work paired organizational leadership with technical execution. The Head of Data and Strategy defined what information the hospital needed, which decisions it had to support, and why those priorities mattered. Momo Analytics translated those requirements into reporting logic, architecture, and a production platform, built by a single engineer end to end.

Six to seven source systems, one pipeline codebase. Clinical and operational data came out of a mix of relational systems and EMR extracts. Each source got its own ingestion job in Python, landing into a common staging layer instead of a patchwork of one-off export scripts.

Every pipeline and transformation defined as code, run through CI/CD. dbt handled the transformation layer, with daily jobs orchestrated automatically rather than triggered by hand. Every change to a pipeline or a model moved through Git before it touched production data.

The whole environment could be torn down and stood back up from a clean slate. Because ingestion, transformation, and the analytical layer feeding dashboards all lived in version-controlled code and config, rebuilding the platform didn’t depend on anyone’s memory of how it was set up the first time. The full environment, source connections included, came back in about an hour, with zero data loss, tested by actually doing it rather than assuming it would work.

Reporting built on tools the hospital could run without a specialist. Tableau powered the dashboards and the hospital’s annual board balanced scorecard, chosen over anything proprietary enough to leave the hospital stuck if the person who built it moved on.

The outcome

The hospital went from spreadsheet-based reporting to a governed analytics platform with a proven, tested disaster recovery time of about an hour and zero data loss on rebuild. The platform fed the annual board scorecard and day-to-day operational reporting without requiring a dedicated engineering hire, and it later became the foundation the hospital’s EMR migration ran on top of.

The result came from a clear partnership: hospital leadership set the direction and kept the work anchored to organizational decisions; Momo Analytics designed and built the systems that made that direction operational. The hospital gained infrastructure that reflected its strategy and could be maintained, inspected, and recovered by the people responsible for it.

Read the next step in the evolution: A Legacy EMR Migration With Zero Data Loss.