Continuous intelligence relies on platforms, architectures, and software that allows organizations to collect, organize, and analyze data to enable fast actions in response to real-time events.
Businesses today need to make fast decisions as events are taking place based on the analysis of streaming data from multiple sources. That’s quite different from the way organizations worked before. And it requires changes in the way data is collected, analyzed, and incorporated into business processes.
A properly implemented approach would allow companies to become proactive rather than reactive. That would represent a radical change in the way businesses operate.
In the past, companies would take action after an event happened. That could include such things as fraud being committed, an operation being shut down due to the failure of a part or equipment, product in the marketplace having a defect, or a service not meeting specifications or customer expectations.
See also: The Case for Continuous Intelligence
Reacting to such events after they happen is bad for business. A company’s reputation is damaged, and there potentially are significant financial implications. Examples of the costs incurred after a problem happens include:
- Criminals engaging in complex identity fraud schemes stole $16.8 billion in the U.S. last year. Financial institutions rarely recoup this amount, and many must pay fines, and for several years’ worth of credit monitoring services for every individual impacted.
- An average offshore oil and gas company experiences about 27 days of unplanned downtime a year, which can amount to $38 million in losses.
- Undetected automotive product defects can incur huge costs, settlement fees, and reputation damage. Several cases illustrate the scope of the problem. They include the recall of 800,000 cars due to an ignition switch problem, 6.5 million tires due to defects, and 100 million airbag inflators. Estimated costs for the recall, injury claims, and settlement fees for these problems were, respectively, $4.1 billion, $5.6 billion, and $24 billion.
What if businesses could detect the problems that led to such fraud, downtime, and defects as they were happening and had taken corrective actions to stop them? That would have saved all that money.
Traditional analytics approaches simply cannot help. In the case of fraud, analysis might be able to help identify outliers to normal transactions. But at best, traditional analysis might help understand what happened after the thief committed the crime.
What’s needed is the ability to examine situational data and use artificial intelligence (AI) and machine learning (ML) algorithms to detect that something is wrong in real time and to suggest or take actions to stop it automatically.
In the case of fraud, such an analysis would allow a financial institution to stop the transaction from taking place.
A similar approach could be used in the case of the downhole well equipment. Data collected from sensors embedded in the equipment could help spot in real time a valve, gasket, O-ring, or bolt that is about to fail. The part could be replaced immediately, thus preventing any unplanned downtime and the associated time to identify the root cause of the problem.
The same holds for the automotive products. Internet of things (IoT) devices could be integrated into a production line to make measurements of temperature, chemical compositions, pressures, and any parameter that impacts quality. Analyzing that streaming data as parts are made and applying AI or ML techniques could help detect defects before hundreds of thousands of products are produced, keeping them from failing when used on the roads.
Enter Continuous Intelligence
The approach described above essentially describes the core capabilities of continuous intelligence. Continuous intelligence relies on platforms, architectures, and software that allows organizations to collect, organize, and analyze data to enable fast actions in response to real-time events.
Unlike traditional analytics, it relies on both the analysis of historical data and real-time data from a variety of sources. That data can include email messages, clickstreams, social media, data logs, and information from sensors and IoT devices. Additionally, CI makes use of AI and ML to enhance further traditional analytics used on such datasets.
Such capabilities give businesses new tools to increase operational efficiencies, reduce downtime, and improve and strengthen customer interactions. Compared to normal analytics and business intelligence (BI), CI offers situational awareness, prescribes actions, and allows businesses to be proactive.
Stepping-Stones to Automation
CI has the potential to change normal business operations radically. Embedding CI into processes and systems lets companies move from reactive organizations, taking actions after an incident has happened, to proactive, taking actions to prevent the incident. So, a business could cut off a fraudulent transaction in progress or fix a problem on an assembly line before defects impact product quality.
That’s just the basic benefit of CI. If a CI effort combines prescriptive analytics with AI and ML, besides knowing something is happening in real time, a solution could suggest what actions need to be taken. As a result, decision-making can be standardized across an entire organization. For example, AI-based rules could be developed based on the knowledge of company experts. If X, Y, and Z conditions occur, take actions A, B, and C. That knowledge is then available, for example, to workers in a remote office who may not necessarily know such real-time events indicate fraudulent activity or production problems are occurring. Used in this manner, CI acts as an expert system providing real-time decision-making support.
Businesses also could take CI a level higher. By knowing something is happening in real time and what actions need to be performed, why not just automate things and have a system to carry out the appropriate actions? Certainly, a business process could be stopped or changed based on real-time analysis of streaming data in a CI solution. And maybe one day soon, a robotic system could change a defective part before it causes an assembly line shutdown.