Oracle Analytics Weekly Roundup — 18-May – Oracle Analytics by Adrian Ward

Below is this week’s Oracle analytics and data roundup for obiee.co.uk readers, covering OAC/OAS, Fusion Data Intelligence, Autonomous Data Warehouse / 23ai, and a practical SQL tip. This edition is intentionally light: there were no specific articles or announcements provided to summarise, so the focus is on stable capabilities, patterns we see in current projects, and how to get more value from what you already have deployed.

Oracle Analytics (OAC & OAS)

Staying Productive When There’s No Big Release

TL;DR: Even in quieter weeks with no major patches or features, you can usually find performance and usability wins in your existing OAC/OAS deployments by revisiting data models, usage tracking, and security design.

No specific Oracle Analytics Cloud (OAC) or Oracle Analytics Server (OAS) release notes or blog posts were provided this week, so instead this section focuses on “evergreen” actions you can take when there’s a lull in new features:

Review usage tracking: If you have usage tracking enabled, use it to identify slow dashboards and analyses. Often, a handful of poorly designed reports account for a disproportionate share of user frustration.
Check logical model complexity: In both OAC and OAS, complex logical table sources and unnecessary snowflaking can lead to inefficient queries. Simplifying joins and consolidating LTS definitions can yield immediate performance gains.
Revisit row-level security: Security filters that are overly complex or implemented at the wrong level (e.g. in multiple subject areas instead of centrally) can slow query compilation and increase maintenance overhead.

Where to look for official updates when they do arrive:

Those documentation hubs are the safest place to confirm what is actually available in your tenant or on-premise environment before you plan upgrades or new features.

Fusion Data Intelligence (FDI)

Conservative Planning Around FDI Capabilities

TL;DR: With no new FDI-specific announcements in the provided material, treat the current public documentation as your single source of truth and avoid assuming roadmap items are available in your environment.

Fusion Data Intelligence (FDI) continues to be an area of interest for Oracle SaaS customers, especially those on Fusion ERP and HCM. However, there were no concrete FDI articles or release notes supplied this week, so it’s important not to infer or assume new capabilities.

Practical steps you can take now:

Validate what’s enabled in your pod: Check which data products and subject areas are actually provisioned and visible. FDI availability can differ by region, product family, and entitlement.
Align with current documentation: Use the official docs to confirm supported pipelines, data domains, and extensibility options before designing downstream OAC or OAS models.
Plan for incremental adoption: Start with a narrow use case (e.g. finance or workforce analytics) and build a repeatable pattern for governance, data refresh, and semantic modelling.

Key reference entry point:

https://docs.oracle.com/en/cloud/saas/index.html

From there, navigate to the Fusion application you use (ERP, HCM, SCM, CX) and then to the analytics and data intelligence sections to see what is formally documented and supported.

Autonomous Data Warehouse / 23ai

Working Within the Documented Feature Set

TL;DR: In the absence of specific 23ai or ADW announcements this week, focus on strengthening your existing Autonomous Data Warehouse setups: resource management, data loading patterns, and cost control.

No new Autonomous Data Warehouse (ADW) or Oracle Database 23ai items were provided, so this section emphasises conservative, documentation-driven practice rather than speculative features.

Areas worth revisiting periodically:

Service-level configuration: Validate that your ADW service size, auto-scaling settings, and workload separation (e.g. multiple services for ETL vs BI) still match your actual usage patterns.
Data loading and transformation: Review whether you are using the most appropriate mechanism (e.g. Data Pump, SQL*Loader, external tables, or tools like Oracle Data Integrator / Data Integration) and whether transformations should happen in the database (ELT) rather than upstream.
Cost and performance monitoring: Use the built-in metrics and AWR-style views to understand which queries consume the most resources and whether indexing, partitioning, or model changes would help.

For authoritative information on what ADW and 23ai currently support, use:

These are the best places to confirm feature availability and syntax before you start using anything you’ve only seen mentioned in presentations or marketing material.

SQL Tip of the Week

Safely Aggregating with Conditional Logic

TL;DR: When you need conditional counts or sums in Oracle SQL, use CASE expressions inside aggregate functions instead of multiple passes over the same table.

A common pattern in analytics projects is to produce multiple conditional metrics from the same fact table. Rather than writing several queries or joining subqueries, you can use CASE within a single aggregation. This is fully supported and well-documented in Oracle Database.

SELECT
    order_status,
    COUNT(*)                                   AS total_orders,
    SUM(CASE WHEN order_amount >= 100 THEN 1 ELSE 0 END) AS high_value_orders,
    SUM(CASE WHEN order_date >= DATE '2024-01-01'
             THEN order_amount
             ELSE 0
        END)                                  AS amount_2024_onwards
FROM   sales_orders
GROUP  BY order_status
ORDER  BY order_status;

What this does, conservatively and without relying on any undocumented behaviour:

COUNT(*) returns the total number of rows per order_status.
SUM(CASE ... END) counts “high value” orders (≥ 100) by summing 1 or 0.
The second SUM(CASE ... END) adds up order_amount only for orders from 2024 onwards, returning 0 for earlier rows.

This pattern is efficient and widely used in OAC/OAS data models and reports, especially when you want multiple derived measures from a single scan of a large fact table.

For more on conditional expressions and aggregation in Oracle SQL, see:

That’s it for this week’s roundup. With no specific items to summarise, the emphasis has been on conservative, documentation-aligned practices you can apply immediately in OAC, OAS, FDI, ADW, and Oracle SQL.

Weekly News