LUNOS on Lakebase: a shorter path from insurance operations to governed intelligence

Executive perspective

Silvermoon has configured LUNOS to run its operational transaction workload on Lakebase Postgres, bringing LUNOS operational data and Databricks analytics into a common Databricks platform architecture while preserving separate transactional and analytical workloads. Lakebase provides a native route for operational change to become governed analytical evidence without a separately operated external CDC stack. Silvermoon is not aware of a published example of an insurance policy administration platform running its OLTP workload on Lakebase.

Silvermoon configured LUNOS to use Lakebase Postgres as its OLTP database. Because Lakebase is native to Databricks, operational change enters Unity Catalog Delta tables through Lakebase Change Data Feed rather than a separately operated external CDC stack. This shortens the route from insurance transaction to governed analysis while preserving distinct operational and analytical workloads.

Why LUNOS makes the path more valuable

The advantage is not proximity alone. LUNOS gives each party, policy, lifecycle event and financial transaction an explicit place and relationship. Downstream analytics begins with insurance meaning already encoded in the operational record, reducing the reconstruction needed before measures can be trusted.

Insurers generate valuable information across customer interactions, quotations, policies, premiums, claims, payments and service activity. Yet this information often spans operational systems, external replication layers, reporting tools and manually maintained processes. By the time it is reconciled and converted into management information, it may have lost both immediacy and insurance context.

This creates more than a reporting problem. Operations see bottlenecks late; finance reconciles competing figures; business teams struggle to trace results to the records and rules that produced them; and technology teams maintain duplicated pipelines, extracts and business logic. Leadership may see several technically correct numbers without a single explanation of how they relate to the customer, policy or underlying event.

A shorter path creates value through four executive lenses

The LUNOS model is deliberately normalized for transactional integrity, reuse and precise relationships. Databricks provides the complementary path: capture operational change, improve quality progressively, govern access and lineage, define reusable measures, and present information through analytics, dashboards or approved operational signals. The two workloads remain accountable for different jobs even when they share one platform.

Bronze, Silver, and Gold make three different business promises

The diagram and evidence fields below use the same five stages. A shorter path is not a shortcut around architecture discipline.

Architecture creates opportunity; a dashboard creates value only when it supports a decision. It must trace back to operational records, refresh to an explicit service level and have an owner when the number is wrong. The work is therefore as much about semantics, controls and accountability as visual design.

The strongest dashboards make the evidence path visible before the visual treatment is celebrated.

Four leadership views of the same governed evidence

Analytical views were created across parties, policies, financial transactions, ledger records and customer relationships. Within the controlled proof environment, AI-assisted query development supported the construction of views and dashboard content; all output was reviewed by Silvermoon Data Specialists before use.

What the proof established

• LUNOS operational data is surfaced in Databricks and transformed into governed dashboards.
• Insurance semantics — including parties, relationships, policies, roles and financial activity — remains traceable.
• Databricks BI capabilities supports Customer 360 insight and actionable lead-generation intelligence.

Silvermoon demonstrated an architecture in which LUNOS operational insurance transactions run on Lakebase Postgres and flow into Databricks analytics through a governed change path. This creates a shorter and more traceable route from policy, party, claim and financial activity to decision-grade dashboards and operational intelligence.

LUNOS is strongest where operational insurance meaning must remain explicit: which policy or claim is involved, who played each role, what changed and how money moved. Each party has one explicit and unique home in LUNOS even when connected to multiple types of agreements (for example a producer agreement) or policies, claims or intermediary relationships; the operational record preserves the context needed to explain every downstream number.

The differentiated architecture goes further than placing an operational platform beside an analytical one. LUNOS runs its transactional workload on Lakebase Postgres; Lakebase Change Data Feed provides a native route for changes to become Unity Catalog Delta evidence for downstream pipelines and audit. Databricks then reconciles and governs that evidence, defines reusable measures and delivers dashboards or approved signals. The advantage is a native route with fewer integration components while governance and control obligations remain explicit.