Partnership: Databricks

Data platform.

Overview

Working alongside Databricks is the discipline of meeting customers on the data platform that increasingly powers their analytics. Operations data and analytics data overlap more every quarter; the integration produces value where they meet. Five surfaces carry most of the weight: Delta Lake as source, Unity Catalog for permissions, SQL warehouse for queries, notebook integration, cross-team alignment.

• Delta Lake as source. Operational metrics and traces read directly from Delta Lake. Existing data-lake patterns extend to operational data.
• Unity Catalog awareness. Permissions flow through Unity Catalog. Access controls preserved across the boundary.
• SQL warehouse queries. Standard SQL access against Databricks warehouses. Analytics on operational data without a separate query engine.
• Notebook integration plus cross-team alignment. Notebooks reference Nova insights inline; ops and data teams work on shared substrate. Real collaboration rather than parallel dashboards.

The approach

Delta Lake first, Unity Catalog aware throughout, SQL-friendly surfaces, documented setup. The discipline is treating Databricks as the shared platform rather than a separate silo to bridge.

• Delta Lake as source. Operational data lands in Delta Lake natively. Matches the customer's existing data-lake patterns.
• Unity Catalog permissions. Access controls flow through Unity Catalog. Security model stays consistent.
• SQL warehouse queries. Standard SQL access matches data-team tooling. No new query language to learn.
• Notebook references plus documented integration. Notebooks link to Nova insights for inline analysis; per-step setup guides support self-service onboarding.

Why this compounds

Operations and data teams work on shared substrate. Historical analytics on operational data unlock investigations the team could not run before. Each integration teaches the team more about how the data platform actually gets used in practice.

• Cross-team collaboration. Ops and data teams share data. Investigations get faster; arguments about which dashboard is right go away.
• Better incident analysis. Historical analytics on operational data. Investigations reach further back than tracing systems alone allow.
• Preserved customer investment. Existing Databricks setup gets leveraged. Adoption costs stay low.
• Year-one investment, year-two habit. First integration is investment-heavy; subsequent customers benefit from the patterns established.