Databricks + ServiceNow CSDM for Service-Health Correlation
A retail bank correlated 18 observability tools to ServiceNow CSDM via Databricks — service-health view that survived an outage drill.
May 9, 2026·2 min read·By admin
At a glance
Industry
BFSI
Topics
csdm · databricks · observability
Published
May 2026
Headline outcome
4x Service-health view validated in outage…
Measured outcomes
What this programme actually delivered.
4x
Service-health view validated in outage drill (resolved 4x faster than baseline)
28 minutes
Mean time to detect: 28 minutes → 4 minutes
-20
Blame-storming time in post-mortems eliminated for top-20 services
6
Pattern in productisation across 6 retail bank products
The challenge
The bank had 18 observability tools (split across IT teams, business teams, and security). ServiceNow CSDM was the system of record for services. Correlation between observability signals and CSDM services was manual; in incidents, blame traveled before clarity did.
What we built
Databricks as the correlation layer. Each observability tool streams to Databricks; CSDM service map streams to Databricks. Correlation rules join signals to services. ServiceNow Event Management consumes the correlated stream. Outcomes flow back to dashboards + business-service health views.
Outcomes
Service-health view validated in outage drill (resolved 4x faster than baseline)
Mean time to detect: 28 minutes → 4 minutes
Blame-storming time in post-mortems eliminated for top-20 services
Pattern in productisation across 6 retail bank products