Autonomous Analytics- Predictions for Business Intelligence in 2030
I recently read an article in Information Week titled: “We’re Entering a New Era of Augmented Analytics.” The author, Jen Underwood, writes: “Next generation augmented analytics capabilities can automatically prepare and cleanse data, perform feature engineering, find key insights and hidden patterns. Automation expedites investigation across millions of variable combinations that would be too time consuming for a human to do manually.” Machine assisted analytics are already implemented in many BI tools, even our own Arcadia Data has “Instant Visuals” which allows the BI tool to make suggestions on the right visual given selected measures and dimensions. I want to take things a bit further. The future is a vast configuration of active, almost aware, algorithms that are constantly looking for trends, outliers, and opportunities. Consider the business intelligence of the future to be almost sentient or human like.
By 2030 business intelligence will transform from a passive tool where humans seek out information to one where information seeks out its consumer through highly advanced bots, what I call ‘Active Seek.’ Converged analytics will take us beyond ‘what happened’ and seamlessly answer ‘why something happened’, and ‘predict what could happen next.’ Lastly these bots will cooperate with one another, forming a vast network of things, ‘Connected Bots’ that control actions of other bots or perhaps even you.
Active Seek: Data Actively Seeks Its Consumer
One of the first things many executives do in the morning is open up their BI dashboard. They review key performance indicators, trends, other visuals that concern their business. Some executives receive email alerts informing them of thresholds that matter with links to take a closer look. To get to this level of sophistication probably took thousands of information technology hours to cleanse, prepare, and develop the tools that these executives use. They are slow to change, expensive, and are very static. In this scenario data and information is passive, waiting for a tool or human to use it. This type of BI tool is just not enough and will be a dinosaur in 2030.
In the year 2030 business intelligence will take on a ‘data-phagocytosis’ state through the use of bots. Much like a physiological process of homeostasis which regulates an organism’s state, ‘Active Data Bots’ will seek out data in a process I call ‘Active Seek.’ Think of ‘Active Bots’ as a collection of white blood cells absorbing bacteria to fend off infection. These bots will be algorithms that will seek out trends, anomalies, and patterns. Via APIs they will integrate their findings into visual tools, alerting tools, control systems, and possibly interrupt human behaviors with new information. They will learn who the right consumer of this information is by following trends and patterns of use. BI vendors will compete to build the best bots that are programmable, flexible, and not super noisy. Just like the volume on your favorite music device you will be able to turn it up or down, or perhaps off. You think email and your smartphone are addictive just wait for the ‘Active Seek’ phenomena.
Converged Analytics: The Need to Know Why and What Will Happen
Data and information are lifeless without use. This includes the various velocities of data: real-time streams and data at rest. Today a business intelligence tool is driven by a user, most visuals tell you ‘what’ happened but not ‘why’ something happened or ‘what could’ happen next. The need to know ‘why’ and ’what will happen’ next is what fed our need for ‘Data Science.’ The business intelligence tools of tomorrow will see a convergence of data science and business intelligence into a single set of tools that will be easy to use and seamless to its consumer or application programming interface (API).
Data science is a very difficult discipline, it requires super humans capable of understanding four things: domain knowledge, mathematics, software programming, and different advanced analytic techniques. The human aspect of this is the problem, as data science takes time, is slow to react to change, and if not properly managed it decays and conforms to the 2nd law of thermodynamics: entropy. However, we need bots that can converge traditional business intelligence and data science to answer the questions organizations need: ‘What happened’, ‘Why it happened’, and ‘What will happen.’
The ‘Active Seek’ bots will evolve and gain the ability to perform and answer all three of these questions. They will use ‘attribution’ analysis to answer questions about why things happened and predictive analytics to make recommendations of what will happen next. ‘What’ happened is still critical because it allows us to focus on the boundaries of time to understand performance as well as measure predictive accuracy as we operationalize predictions. BI vendors will compete over who has the best converged analytics in the year 2030.
Connected Bots: Neural Network of Things
Today’s business intelligence is mostly monolithic. It lives outside of the working enterprise as a disconnected source of information detailing organizational performance. It works within human processes that hope to operationalize change in business activities in order to optimize revenues or decrease costs. Don’t get me wrong, these things are great. However, they are not enough. We need business intelligence to be connected to our world.
If we can learn anything from the Connected Vehicle initiative it is the need for everything to be connected to everything. Tomorrow’s Connected Vehicle will be connected to other vehicles, road signs, people, and traffic signals; it is a massive collection of sensors and streams of data. IoT sensors already generate trillions of bits of information every second but there is one thing missing? How do we consume all these streams? Today we see organizations like Confluent, the inventor of KSQL. KSQL is a technology which will democratize streaming analytics so that they can be exploited by business users and connected bots. IoT has more to do with algorithms that interpret data and decide what to do next in a complex array of autonomous bots connected through APIs.
By the year 2030 the amount of sensors and data flows will require algorithms, or bots, to be able to take advantage of these capabilities. Not only will these bots be able to analyze data in real-time but they will also be able to communicate detailed information much like the human brain, through APIs. Bot APIs will act like dendrites and synapses enabling communication and potential actions. The collection of bots are the real power behind IoT relaying vital information in a homeostatic means of self preservation and healing. These bots will be always active, always working; they will bring together descriptive and predictive analytics as to understand why something happened and what will happen next. Business intelligence and visualizations will have to work with these bot collections in order to make them consumable by people. The adoption of this technology will be the biggest technical revolution of our existence!