Žiga Vaupot's Blog

In my previous blog posts, we explored this fascinating new technology— how to set it up , the broader analytics paradigm it introduces, and a deep dive into some specific Oracle Analytics functions. But what about the basics? What about standard business reporting—the business-as-usual activities we deal with every day? For example, the classic financial reports that managers rely on daily to run the business. These are the standard reports that are typically pre-prepared and embedded within dashboards. Well, let's take a look! I have tested two examples of those standard financial reports : Revenue analysis report that compares current year actuals vs. previous year, vs. planned data and calculates year-to-date values. Performance report that is providing information about revenue, profit and profit margin for the selected period, ie. half-year. The process is already known by now. Startup Claude client and start asking questions. When working with OAC MCP Server, ...

This is summary blog post with references to posts published under Oracle Analytics Semantic Modeler theme: Introduction 1. Creating a new System Connection 2. Creating a new Semantic Model & Connections 3. Physical Layer 4. Logical Layer 5. Presentation Layer 6. Calculations 7. Variables 8: Using Data Filters to enable Row-Level Security 9. Data Fragmentation 10. Setting up Git repository

This is summary blog post with references to posts published under Unlocking Oracle Analytics Cloud with AI theme: Oracle Analytics Cloud MCP Server Facing a new paradigm in business analytics Going deep dive on OAC MCP Server Standard Financial Reporting using OAC MCP

When I first tried the Oracle Analytics AI Assistant, I was genuinely surprised by what it could do. Most queries ran very well—even in Slovenian, a language not officially supported in Oracle Analytics ☹️—and I was able to truly talk to my data. However, when I tried to push the AI Assistant toward more advanced analytics, such as applying functions like TRENDLINE or CLUSTERS —both of which are one-click advanced analytics features—it simply didn’t understand what I was asking. When the Oracle Analytics Cloud MCP Server became available, I decided to see whether it could handle these advanced analytics functions. And it did. TRENDLINE First, I asked the MCP Server to analyze sales revenue for the last two years of available data . Since my dataset is fairly old, I wanted to avoid any misunderstandings, so I explicitly specified that the analysis should focus on the last two years of available data. So, here’s how it all went: Click here to ...

Introduction A new paradigm is reshaping business analytics. Instead of the traditional build a dashboard, then interpret it approach, organizations are moving toward interactive narrative analytics; a model where users begin with natural-language questions, the system retrieves governed and trusted facts through retrieval-augmented generation (RAG), and large language models (LLMs) synthesize explanations that users can probe, refine, and challenge. The experience evolves from static charts and filters into a continuous analytical dialogue: what happened, why it happened, what may happen next, and what actions to take—all anchored in a shared semantic layer and consistent metric definitions. In this shift, LLMs redefine the user interface , RAG redefines trust , and MCP servers redefine integration . RAG constrains generative models to validated enterprise sources—datasets, KPIs, subject areas, and documentat...