Even with the best intentions, people can't possibly analyze all of the data at their disposal. Those that have a plan, however, will see a great reward over their competitors; a $430 billion advantage, according to projections by analyst firm IDC.
What’s needed is a way for enterprises to generate business insight at breakneck speed that applies next-generation thinking across its entire portfolio to help lower cost, reduce risk, accelerate innovation, and get predictive insights. And by these standards, we suggest Oracle Autonomous Analytics.
Oracle President of Product Development Thomas Kurian recently demonstrated the latest advances in the Oracle Cloud Platform. Among the many self-driving, self-securing and self-repairing Cloud Platform services, Kurian demonstrated how Oracle Analytics will include automated data discovery and preparation as well as automated analysis for key findings along with visualization and narration delivering quicker insights.
The multiple autonomous database services, each tuned to a specific workload, are expected to be available in upcoming versions, including Oracle Autonomous Data Warehouse Cloud Service for analytics.
What do we mean by Autonomous Analytics? Not artificial intelligence and not simply automated, it is the application of machine learning to human judgment. This new collaboration creates many opportunities to improve your decision-making and predictive abilities as the data landscape becomes more and more complex.
Machine learning is key to enabling Autonomous Analytics. Machine learning is not, however, about writing more rules. A systemic problem today is that most predictive systems are built on millions of lines of code with nested business rules that are complicated and expensive to maintain. Never mind launching an audit. Instead, Autonomous Analytics refers to self-learning algorithms that thrive with the growing volume of data. Unlike prior generations, these algorithms modify themselves as more data and more actions are evaluated.
Adaptive and Autonomous
The future of business analytics is both adaptive and autonomous. This means using machine learning to power the business analytics value chain which starts with discovery, moves to preparing and augmenting data, then to analysis, modeling and finally to prediction. It is data-driven and is a powerful Platform for innovation. Personal context is all-important: the system must understand who I am, where I am and what I need to know now.
The objective of data discovery and preparation stages is to help you easily find useful datasets across any combination of sources. It means understanding which datasets you can access and what condition they are in with regards to quality and completeness. Then to receive automated recommendations for data standardization, cleansing, and enrichment.
Rather than starting your visual analysis with a blank canvas, you can now start with insights that are automatically generated, based on correlations and patterns in the data that the system identifies. This autonomous capability speeds up analysis provides automatically generated narration and delivers quick real-time insight.
Pictured above: Oracle Day By Day
Today, on your mobile device, you can verbally ask business questions (natural language processing) using your own business- and company-specific vocabulary, have the system understand your requests and anticipate your questions based on self-learning. It will reply to your voice queries with multiple possible answers, then learn and refine based on your guidance.
Currently, data-savvy executives can create visual business models without any special training, collaborate with others on the models, then test assumptions without impacting anyone else, in order to see answers to "what happened" and "what if". But now we can create a new value when these models automatically learn from transactions and update predictive results in real time. Not only does this reduce administration overhead, but it also eliminates human bias. This allows real-time learning to be immediately available for the next prediction, for example, to drive adaptive, high-value interactions with customers, to determine the best profile for employee candidates, or to recommend when to service critical machinery and capital assets.
For business professionals, Autonomous Analytics can provide insight to inform every decision, when and where it matters, and in context. For IT, automated insights result in lower cost and enable business users to be more self-sufficient, lowering the support burden.
Autonomous capability, driven by machine learning, is a promising enabler of greater productivity and innovation.
How would you feel about an autonomous approach to your data? Feel free to leave a comment below.