Is your organization ready for AI? Sort through the hype and find the most practical solution

It seems as if everywhere you turn, someone’s referencing AI. “Artificial intelligence is essential for business,” states a news article. “AI is the only way forward,” insists a colleague. And the scariest of all: “If you don’t implement AI quickly, you’ll be left in your competitors’ dust.” The constant refrain of AI, AI, AI can leave you feeling like your organization has lost the race before it has even entered it.

The truth is, AI is changing the world at a remarkable pace. And, eventually, nearly every industry and business will benefit from it. “Whether you work in retail, banking, transport or the public sector, AI will be an integral part of the way you do business in the future as it has huge potential to improve decision-making, increase efficiency and power new ways of working,” states the article How to Get Your Business AI-Ready.

But that doesn’t mean AI is the best solution for your organization right now. Implementing AI takes a massive commitment in the form of time, resources, and money. It requires a critical mass of data and properly trained staff. By prematurely jumping into a high-profile AI program, you risk ignoring the valuable tools already available and stalling other strategic projects underway. Instead, a more practical approach—one that uses software scaled to your operations—will move the most important metrics now, while developing an analytics culture that will make AI more feasible down the road.

AI has become a buzzword. What exactly does it mean?

In its prolific use, the meaning of artificial intelligence has become skewed. Some have come to view it as a magical solution capable of instantly transforming business all on its own. Others equate it with automation. Neither of these is true, however. AI requires much preparation and strategy (more on that later), and where automation follows pre-programmed rules, AI involves machine learning. AI is designed to mimic human thinking by making predictions and adjusting its processes based on new data insights. In this way, AI is quite different from many previous technological revolutions, during which technology took over specific, static roles within processes.

In his article Is Your Business AI ready?, Tariq Mustafa summarizes AI this way: “[AI’s] capabilities revolve around grouping things based on similarities, predicting new data points based on previously available information, detecting anomalies in otherwise fixed patter[n]s of behaviors and suggesting things that are likely to be done next.”

The foundational elements your organization needs before implementing AI

AI has certainly made an incredible difference for many organizations. One example is Netflix’s “suggestions” feature. The video-streaming service uses AI to continually collect data points (which shows you watch, for example) and generate suggested material based on that aggregated data. Consumer sites like Amazon employ AI in a similar way, using customer searching and purchasing data to suggest like products.

Although these examples may seem unremarkable, they are incredibly complex. To implement AI, organizations first need several essential components.

Foundational element #1: Lots of data, and the right data

Data is key to making AI work. Tariq Mustafa writes, “If the business is not collecting the base data in the first place – let’s say due to the off-line nature of the process or lack of digitization of the process, even the best AI and [machine learning] facilities will not be able to do any intended good.” But even if collection is occurring, the amount and quality of the data matters, too. For example, if an organization is processing a few dozen claims a year, AI will not provide critical insights into claims management. AI works so well for behemoth companies like Netflix and Amazon because they have access to immense amounts of data, which is necessary for fueling the AI machine. To fully unleash the power of AI, you must first achieve critical mass from a data perspective.

Before even considering AI, Mustafa suggests that businesses “revamp and re-engineer their entire process chains with an added emphasis of collecting digital data at as many points as possible and viable.” This helps ensure that your organization is not only collecting valuable data, but also being mindful of the processes used to gather it. Flawed processes and data plus AI still equals flawed processes and data.

Foundational element #2: A team of highly analytical people

The CMS Wire article 5 Signs Your Company Isn’t Ready for AI states, “Trying to deploy AI without the right team practically guarantees failure.”

For AI to be effective, you need a team of people skilled in strategic thinking. They must know how to collect and analyze information, and troubleshoot when necessary. They must have a laser-sharp focus on specific goals. (This means an ambiguous objective like “make things more efficient” won’t cut it.) AI success comes as a result of extensive trial and error, and an ability to understand where to correct mistakes or shift direction.

Foundational Element #3: A culture of change and willingness to experiment

A Genpact/Fortune Knowledge Group survey determined that 50% of senior executives believe one of the most important staff qualities for adopting AI is “the ability to adapt to change.”

How does one create an adaptable culture? By getting staff comfortable with experimentation, a highly valued concept in start-up culture. Leadership can encourage this by assuring staff that failure, rather than resulting in negative consequences, is a valuable tool for progress. “With any emerging technology, you need to expect some stumbles and pitfalls along the way,” states the CMS Wire article. “It’s inevitable. Companies that succeed in the AI revolution will set themselves apart because they will recover from failures more quickly (and because they allowed themselves to have failures in the first place).”

The strength that comes with success after failure will ready staff to take on the ups and downs of implementing AI later on.

How you can move metrics with the data and tools you have

Implementing AI is a beast of a process, even for businesses that are fully equipped and prepared. Doing so requires not only time and resources, but also the re-engineering of internal processes.

By beginning with the suggestions that follow, your organization will have the essential building blocks in place to make AI a reality in the future, and also be able to make major improvements in KPIs today. Whether you’re expected to implement an enterprise risk management (ERM) program, improve total cost of risk (TCOR), or streamline your claims management process, the below functionalities will help lead to fast, positive results that tie to your most pressing strategic objectives.

Automated processes

Start by doing an evaluation of your organization’s systems and processes. Consider where your organization’s highest priority problem areas are. What is costing unnecessary money? Where could processes be made more efficient?

The right risk management information system (RMIS) is designed to implement automation where it makes sense. The Harvard Business Review article If Your Company Isn’t Good at Analytics, It’s Not Ready for AI states, “Companies need to automate repetitive processes involving substantial amounts of data — especially in areas where intelligence from analytics or speed would be an advantage.” Origami’s automation functionality takes over previously manual tasks, such as filing reports and sending emails, and sends alerts, tasks, and messages to the appropriate parties in a process.

The efficiencies gained through automation will allow for more streamlined internal communication, saved money, fewer errors, and cleaner data that can then be used strategically.

Data analytics

Data analytics, available in several forms in Origami Risk, takes data and turns it into invaluable insight.

Origami’s Official Disability Guidelines (ODG) integration, for example, provides access to tens of millions of cases, plus decades of research by physicians and methodologists, all of which can be used for benchmarking an organization’s workers’ comp claims against the entire industry and making better predictions about lost time and employee recovery. These predictive assessments offer one of the major benefits of AI without the hassle of AI implementation.

Another example of integrated analytics is Origami’s dashboard functionality built around event-based triggers. Bringing personalization to the claims process summarizes event-based triggers: “Once any object in the system hits certain conditions (predetermined by your organization as problematic), Origami automatically sends notifications, assigns tasks, or generates reports to make sure any claim falling outside the pre-set parameters gets the extra attention it warrants.”

Data insight like this will help push your organization to the next strategic level without slamming your staff with the burdens associated with AI. The article RMIS tricks to avoid your own Groundhog Day states:

The real power embedded in analytics lies not in the data itself, but rather in its ability to trigger the right conversations. This is where real change is born. A RMIS that helps you cut through the noise and spark those strategic, data-based discussions is a key part of turning analytics into more than a buzzword.

So, regardless of whether you eventually head down the AI path, tackle these two items first. The Harvard Business Review article puts it bluntly: “Companies that rush into sophisticated artificial intelligence before reaching a critical mass of automated processes and structured analytics can end up paralyzed.”

One last thing: Be wary of “toy” applications

As use of the term AI continues to spread, the Emerj article Enterprise Adoption of Artificial Intelligence – When it Does and Doesn’t Make Sense warns of taking on technology simply because it uses AI, not because it solves a specific problem for your organization. “Coming up with ways to use AI because other people are doing it makes it a ‘toy’ application because it is not really the best use of your resources,” the article states. “It takes a lot of time, training, and tinkering as well as specialized skills, and often goes nowhere.”

Likewise, organizations should be leery of software with AI that hasn’t yet proven its worth. The Emerj article says that vendors capitalize on organizations’ interest in AI “because they need guinea pigs to ‘pilot’ products, and they’ll sometimes encourage closing deals even if they are not well organized…You know this is happening when the question becomes ‘What can we do with AI?’ rather than ‘How can we best use our resources to meet our business goals?’”

Finally, know exactly what types of solutions you’re getting from your software, and what they’re worth. Akhil Talwar, senior product lead for Bold360 by LogMeIn, told VentureBeat, “The key thing is businesses have to be sure they’re not paying AI prices for something that’s basic automation, or might look like it’s more advanced than it actually is. Those are some of the things that really make an impact from a business decision standpoint — making sure you have the right tools and technologies in place to deliver on what they’re looking for.”

Origami Risk offers a secure, cloud-based infrastructure with automation and data analytics functionalities that work with your existing data, scale of operations, and staff to move the needle in big ways.

Contact Origami Risk to see how we can help you move the metrics that are important to you right now, with the technology you have in place right now.