Oracle Autonomous AI Lakehouse 26ai: Embedded Essbase Is Here

We are thrilled to share that Oracle Autonomous AI Lakehouse 26ai is here, and the news our team is most excited about is the embedded Essbase. Multidimensional analytics, planning, and AI-assisted querying now live right next to your governed lakehouse data, on one managed platform.

For finance and analytics teams, that changes the conversation. The lakehouse becomes more than a place where data lands. It becomes a place where the business can slice, model, forecast, and ask questions of that data with the tools they already know.

Why Essbase inside the Lakehouse is a big deal

Oracle frames the lakehouse as open data plus governance, and its KScope26 reflections point to Apache Iceberg support, cataloging, semantic context, and access that avoids copying data into new silos. Essbase brings the other half of the story, which is the business model on top of that data.

With Essbase on Autonomous AI Lakehouse, you keep the dimensional modeling, financial intelligence, and write-back that planners and analysts depend on. You also keep the familiar tooling: Smart View, MDX, calc scripts, and REST APIs. What you shed is operational overhead, and what you gain is tighter alignment with governed enterprise data and a clear path into Oracle's broader analytics and AI ecosystem.

In plain terms, fewer moving parts, fewer copies of the data, and one platform to secure and manage.

The Essbase features that matter most

Essbase has earned its reputation over decades, and every one of its core strengths comes along for the ride.

·       Multidimensional modeling. Hierarchies, dimensions, and members that mirror how the business actually thinks about products, regions, time, and scenarios.

·       Fast calculation and aggregation. Calc scripts and dynamic calcs that turn detailed data into answers in seconds, even across large cubes.

·       Write-back and what-if analysis. Planners can change a driver and see the impact immediately, which a read-only reporting layer cannot do.

·       Smart View and Excel. Analysts stay in the spreadsheet they love while working against a governed, shared model.

·       Open interfaces. MDX and REST APIs make it simple to integrate Essbase into existing applications and automation.

·       Federated data access. Existing semantic models and calculations can work with lakehouse data without forcing a copy into yet another store.

·       Enterprise security. Role-based access carries through to every query, whether it comes from a person or an AI assistant.

AI that makes lakehouse analytics easier

This is where it gets really fun. Essbase now ships a growing set of AI features (the feature names below are Oracle's, and the details come from the Oracle sources linked at the end), and with the lakehouse as the source, they have rich, governed data to work with.

·       AI Query. Ask a question in plain English, like the example in Oracle's own AI Query announcement, "Show me sales by region for Q1," and Essbase generates the MDX and returns report-ready results. You can view, copy, and explain the generated MDX, open results in Smart View, and save them as reports. Vectorizing the cube outline lets the AI map your words to the right dimensions and members.

·       Calculation Assistant. Helps developers write and explain calc scripts, which shortens the path from business logic to working calculations.

·       Ask Essbase. A conversational way to explore the model and its data.

·       Essbase MCP Server. Approved AI clients, including Claude, can discover and call authorized tools to explore outlines, query data, and run calculations, all under Essbase authentication and access profiles.

·       One gateway across Essbase and the Lakehouse. The Oracle Data Studio MCP Server spans Essbase, Data Studio on the lakehouse, and Data Transforms. Its routing tools can look at a question and a fact table and recommend whether a table, an analytic view, or an Essbase cube is the right engine to answer it.

The common thread is governance. AI works through the hierarchies, calculations, and security of your Essbase model, so answers stay consistent with how your business defines its numbers.

As with any generative AI, we recommend validating results before acting on them, and the generated MDX makes that review easy.

Built with Oracle: our early access experience

At iArch Solutions, we were fortunate to have early access to Essbase on Oracle Autonomous AI Lakehouse 26ai. We did not just kick the tires. We worked collaboratively with the Oracle team, putting the offering through real world scenarios and sharing candid feedback on what worked, what did not, and what customers will need.

That feedback helped refine the offering before launch, and it is exactly how we like to work: hands on, practical, and focused on what delivers value in a customer's environment. We are proud to have played a part, and grateful to Oracle for the open partnership.

Stay tuned for our hands-on findings

This article covers what's new. Our next one will cover what it's like in practice. Our team will be testing and deploying Essbase on Oracle Autonomous AI Lakehouse 26ai, and we'll share what we learn: how setup and migration go, how the AI features perform with real models and data, and the practical tips we pick up along the way.

If you're weighing a move to the lakehouse, that follow-up will give you an honest, experience-based look before you commit. Check back soon so you don't miss it.

What this means for you

If you already run Essbase, this is your path to a fully managed platform without giving up the models, calculations, and Smart View workflows your team relies on. If you've invested in a lakehouse and want finance and planning to work directly on that data, embedded Essbase gives you a proven engine with modern AI built in.

Either way, the goal is the same: less time maintaining infrastructure, fewer copies of your data, and faster answers your leadership can trust.

Ready to see what embedded Essbase can do for your team? Schedule a discovery call today.

Because we helped test and refine this offering alongside Oracle before launch, we know what it takes to get it right in a real environment. Whether you're evaluating the platform, planning a migration, or want to see it working against your own data with a proof of concept, we'll help you move forward with confidence.