As cloud adoption accelerates, so does the complexity of managing cloud costs. Without proper visibility and automation, teams risk overspending, budget overruns, and missed optimization opportunities.

Oracle Cloud Infrastructure (OCI) offers detailed cost and usage data through its FinOps standard FOCUS reports. However, simply having access to this data is not enough. What if you could automatically surface anomalies, correlate costs with budgets, and interact with your reports using natural language—all without building complex infrastructure?

This blog series demonstrates exactly that. Using Oracle Autonomous Data Warehouse (ADW), Oracle Analytics Cloud (OAC), and built-in machine learning (OML4SQL), we show how to build a cost intelligence platform that brings AI and automation to cloud financial operations.

About This Blog Series

This is the first post in a 5-part series focused on building a cost analysis and anomaly detection solution in OCI using only Oracle native services—no external compute or third-party tools required.

Part 1: (this post): Solution overview and setup
Part 2: Copy FOCUS reports from Object Storage to ADW using cloud-native DB pipelines.
Part 3: Create ADW schema, ingest budgets data and build visual dashboards using Oracle Analytics Cloud.
Part 4: Train and run an in-database anomaly detection model using OML4SQL.
Part 5: Build an AI agent that allows users to query cost data using natural language.

Prerequisite: Export FOCUS Cost Reports to Object Storage

Before you begin, ensure that FOCUS cost and usage reports are automatically delivered to an Object Storage bucket in your tenancy. Below blog post has detailed guide for setting this up

Automating the Export of OCI FinOps Open Cost and Usage Specification (FOCUS) Reports to Object Storage

This step is essential. The rest of the solution assumes that cost reports are regularly deposited into a known bucket.

 

High-Level Architecture

 

The architecture includes:

  • OCI Object Storage for storing FOCUS cost reports.
  • Oracle DBMS_CLOUD to ingest data into ADW.
  • ADW schema designed for cost and budget reporting.
  • Oracle Analytics Cloud for dashboard visualizations.
  • In-database ML via OML4SQL for anomaly detection.
  • Oracle AI Agents to query data using natural language.

Why This Solution Matters

OCI customers often span multiple Tenancies, regions, compartments and services. Manual analysis of CSV-based cost data is inefficient, error-prone and not scalable. This solution helps teams:

  • Automate detection of unusual spend patterns.
  • Discover misconfigurations or underutilized resources quickly.
  • Enable finance and engineering teams to collaborate using dashboards.
  • Improve forecasting and chargeback models across departments.

All of this is done using native Oracle services, keeping your data secure within OCI and simplifying operations.

What Makes This Solution Different

Here are a few ways our approach stands out:

  • Cloud-native ingestion: FOCUS reports are pulled directly from Object Storage into ADW using built-in data pipelines. No external servers or scripts.
  • Real-time AI insights: Anomaly detection is performed directly inside ADW using SQL-based ML. No ML expertise is required.
  • Integrated dashboards: Visualizations are created in OAC with drill-down support for compartment, service, and budget-level tracking.
  • Conversational AI agent: A natural language interface allows you to ask questions like “What was the highest cost service last week?” and get SQL-backed answers instantly.

In Part 2, we will define the ADW schema, ingest additional budget data and connect Oracle Analytics Cloud to create interactive dashboards. This will set the stage for correlating budget targets with actual cloud spend.