Autonomous Data Warehouse: New innovations for data analysts, citizen data scientists and LOB developers

March 17, 2021 | 8 minute read
Text Size 100%:

Guest post by George Lumpkin, Vice President, Product Management for Autonomous Data Warehouse

After the introduction of Autonomous Data Warehouse, organizations of all sizes recognized how simple it could be to provision a data warehouse. Since Autonomous Data Warehouse requires no operational administration (and thus does not require a database administrator), a cloud data warehouse is within the reach of many more organizations than before. For many IT organizations, Autonomous Data Warehouse was exactly what they needed. Autonomous Data Warehouse addressed the day-to-day operational tasks, so that IT specialists had more time and energy to devote to adding new datasets and extending the data warehouse to meet the business team's requirements.

While Autonomous Data Warehouse eliminated the need for an operational administrator, there was still a need for strong IT and development skills to build a warehouse. Just to start, organizations needed to be able to design their database schema, create their database tables using SQL, load their data sets into tables, and cleanse and transform their data. Autonomous Data Warehouse (and essentially every other cloud data warehouse vendor) was a SQL-oriented solution, designed for technical users who were at ease with relational databases.

Evolving into a Self-Service Model for Data Warehousing

At Oracle, as we started to consider the next generation of cloud data warehouses, we recognized that there was no reason why building a cloud data warehouse should require deep IT expertise. Data warehouses are built to solve business problems -- providing deeper insights about an organization's business data -- and the primary users of the data warehouse are not SQL gurus, but data analysts and other business professionals in the lines of business organizations.

In this blog, we introduce a new focus for Autonomous Data Warehouse: empowering the next-generation data analysts with a next-generation cloud data warehouse. Autonomous Data Warehouse now delivers a suite of built-in, self-service tools designed not for the SQL user, but for the business user, citizen data scientist and citizen developer.

Autonomous Data Warehouse has added a palette of new built-in tools (in addition to the pre-existing built-in tools such as SQL Worksheet, Machine Learning Notebooks and APEX):

  • Data Load
  • Data Transform
  • Business Model
  • Data Insights
  • Catalog
  • AutoML
  • Graph Studio

Self-service data warehouse tools: an example

These new tools are designed for business users, rather than technical users. The starting point for any data warehouse is to load data. Previously, Autonomous Data Warehouse’s only technique for loading data from cloud object storage was via SQL:

If you are a SQL user, then the above code snippets are simple to use. However, if you are a data analyst, you probably do not know SQL … and you are likely not eager to learn these database details. The new built-in tools for Autonomous Data Warehouse have the data analyst in mind, so that data loading becomes much simpler via a UI:

A data analyst could use the UI to simply load data from a local file on their laptop. They would simply click ‘next’ on the above screen and then they only need to drag and drop their data file onto the Data Load tool:

Autonomous Data Warehouse will automatically interpret the files, set up the SQL table definitions with the appropriate column names and data types, and load the data. The user can optionally check the settings before the data load starts, and we can see that the details of the table definition and the options for data loading:

But in the end, data loading has been reduced to essentially a one-step process: drag the data file into the data load tool. Autonomous Data Warehouse takes care of the technical details.

This simple example illustrates how Autonomous Data Warehouse is extending its vision to data analysts, in addition to SQL-oriented technical users. With similar simple tools, Autonomous Data Warehouse provides built-in tools to help the data analysts transform the data within their data warehouse, organize their data into business models (to provide a business-centric view of the data), and search their data with a catalog (which additionally provides data lineage so that business users can understand where their data came from).

Citizen Data Scientists - Introducing AutoML

Autonomous Data Warehouse helps other business professionals beyond data analysts – it also introduces new capabilities for ‘citizen data scientists.’

A citizen data scientist is a business user who has a deep understanding of the data and the business problems that need to be solved – but is not a professional data scientist with an advanced computer science degree.

Autonomous Data Warehouse includes over 30 built-in machine algorithms, a rich set of capabilities that have been built into Oracle’s database over the past 20 years.

The challenge here is that the above algorithms are used by trained data scientists who understand all the nuances of machine learning.

Autonomous Data Warehouse’s new AutoML feature now bridges the gap, allows the non-expert citizen data scientist to create their own machine learning models to tackle business problems such as identifying most promising marketing or sales prospects.

AutoML assists the citizen data scientist by selecting the most appropriate algorithms and parameters based on the provided data – without any special knowledge of the algorithms themselves required.

Here AutoML decides which algorithm is best matched to the specific business problem – in the above example, AutoML has tested a range of algorithms and finally selected a General Linear Model as the best algorithm to solve this specific customer attrition/churn requirement. This truly brings the power to work with machine learning to a much wider business audience.

Want to learn more? Read this in-depth blog post by our machine learning team

LOB Developers - Introducing APEX for Low Code Development

Oracle APEX, Oracle’s low-code development platform, provides you with all the tools you need to build apps in a single, extensible platform, which runs as a part of Oracle Autonomous Data Warehouse.

Using APEX, LOB developers can quickly develop compelling apps by simply combining pre-built, extensible widgets. It’s incredibly easy and fast to deploy rich, powerful, mobile-friendly HTML apps that solve real problems and create immediate value. 

It’s vast array of technologies mean that you don’t have to be a hard-core developer. This means you can focus on solving the problem and let APEX take care of the rest.

To explore how APEX helps you eliminate complexity, deliver results faster use the following to our APEX landing page (apex.oracle.com) where you will find everything you need to get started.

Graph Analytics

Autonomous Data Warehouse continues to expand the breadth of its analytic offerings, with new graph analytics capabilities that allow users to understand and analyze relationships within data. Graph analytics can help business users to understand how their customers interact and identify which customers are most important and influential, or can help to identify fraud by deeply analyzing chains of financial transactions

Autonomous Data warehouse introduces Graph Studio, a built-in low-code interface which automates graph modeling and data management, and simplifies graph analysis and visualization. Importantly, Graph Studio guides the data analyst in creating the graph model. The data warehouse may contain tables with customer data, and Graph Studio can example the customer data, understand the pre-existing relationships present in the database, and create graphs by identifying the vertices and edges based upon those relationships.

Once the graph is created, Graph Studio provides a built-in notebook with graph visualization capabilities:

 

Conclusion

Autonomous Data Warehouse is leading cloud data warehouses in a new direction. The data warehouse does not need to be solely the domain of SQL experts and IT professionals, nor should the data warehouse require the assembly of numerous pieces and components. Autonomous Data Warehouse has expanded the cloud data warehouse service to encompass data loading and transformation, simplify deep analytics like graph and machine-learning and quickly deliver new low-code applications – all with built-in self-service tools.

Start Using These New Tools - Right Now!

This was just a brief introduction to the new self-service tools, but look out for more blog posts covering these various new capabilities. Or, better yet, connect to your own Autonomous Database and try out the new tools for yourself! If you were wondering then YES these tools are included in the Always-Free version of Autonomous Data Warehouse.

Click here to learn more about Autonomous Data Warehouse, or if you’d like to try it for yourself, visit Autonomous Data Warehouse Get Started Page

Guest Author

Oracle Chatbot
Disconnected