artificial intelligence

Artificial intelligence coupled with big data frameworks could yield great results for businesses today. Source: Shutterstock

Rethinking the artificial intelligence factor in businesses today

ARTIFICIAL INTELLIGENCE is being featured in more and more business models today than ever before—they’re no longer isolated to the fringes of today’s business world but are being implemented in many industries as an alternative to huge teams and the inevitability of human error.

Lots of businesses are looking into it. According to KPMG’s 2017 Global CEO outlook report, more than 50 percent of all CEOs are sensing some kind of concern about how their business can integrate artificial intelligence into their basic automated processes.

Evidently, many executives are interested to see how far artificial intelligence can take their business, but is artificial intelligence really a feasible technology to pair with the everyday business?

team meeting

Pre-AI, data management tasks could be extremely laborious, requiring large teams of people. Source: Shutterstock

It’s difficult kind of software to handle, requiring experience with complex data and computer algorithms, skills you probably wouldn’t find in your run-of-the-mill enterprise. Some critics are already calling the hype around the tech a bust, citing a lack of maturity in recent innovations as well a general misunderstanding about the capabilities of such tech in by the decision makers in businesses.

Many consider the technology a silver bullet that is able to magically solve many issues, but artificial intelligence (AI), as with any other tool, has its limitations, especially at this early stage.

SEE ALSO: Artificial intelligence – long way to go from hype to hope

However, it’s worth thinking twice about artificial intelligence before dismissing it outright. The technology may be difficult to understand for the average joe, but in actual fact it’s all about thinking about the parameters in which AI can bring benefits to your enterprise. According to Anil Chakravarthy, the chief executive of Informatica, a cloud and data services firm, AI-led data management could be a significant game-changer for businesses managing massives troves of information.

Anil Chakravarthy, CEO, Informatica. Source: Informatica

“Most enterprises have thousands of databases, data files, applications, and analytics systems,” he said in an email interview with Tech Wire Asia. He explained that AI can be super effective as a tool to harvest and analyze a business’s metadata, which can provide “enterprise-wide visibility” for a “truly meaningful and positive impact on data management productivity.”

If AI is the cool kid in the classroom today, then data is its indispensable sidekick. Data is quickly becoming one of the most valuable commodities in the world today, and companies have tonnes of it lying around without purpose.

“Rather than something that a business simply generates, data must be viewed as an asset that is discoverable and usable by any user across the entire organization,” he said, adding that the data can’t be just whatever is lying around but it must be “of a quality that is fit for purpose: high quality for important decisions and interactions and fair quality for rapid innovation and iteration.”

Businesses can glean rich insights into their performance and areas for improvement if they can find a way to harness that information effectively and accurately.

SEE ALSO: How artificial intelligence is taking Asia by storm

AI offers this capability as it’s able to help businesses catalog and analyze their technical, business, operational and usage metadata across both physical and cloud domains, without as much labor as doing it manually would require. AI can help automate data analytic tasks so that employees are free to explore other work that requires their attention.

Chakravarthy makes it clear though that the introduction of AI isn’t necessarily going to make a company’s in-house data scientists redundant, though. In fact, having these tools will further aid them in their work by making it a faster process, leaving the humans to deal with other processes down the line, such as bringing such data into context and visualizing it for third party users.

data information paper piles

Huge troves of data can offer great actionable insights for companies and AI can help automate many analytics processes. Source: Shutterstock

“Companies are increasingly interested in speeding up how fast they can deliver data for critical business initiatives,” he said, saying that it’s integral that businesses have a data management competency as a central pillar of their philosophies in order to stay competitive.

“While this has traditionally been a simple matter of productivity and manpower, more are turning towards automation. That is where machine learning comes in.”

SEE ALSO: Trend Micro extends AI power to domestic antivirus

So how can we think about the benefits that AI can bring to a business’s data? According to Chakravarthy, there are three thing to consider when thinking about AI’s role in your company: speed, accuracy and automation. They’re all interlinked together, but essentially AI-enabled data management systems are really effective at enabling fast delivery of data to both the business and their clients, as well as alleviating employees of burdensome, painstaking work.

“When it comes to unstructured files, a lot of manual work can be required to deduce what data belongs in what category,” Chakravarthy explained, noting that though humans are good at recognising forms of data, conducting that task at scale is tough.  

“The issue is that scaling human intelligence on unstructured data is a highly manual, and highly time intensive process.”

He said that well-taught AI systems are able to recognize data much in the same way humans do, and automatically sift them into the correct categories and context.

“When faced with huge data sets, it can become difficult to manually obtain an overall picture of all the relationships and patterns available. Similarly, when working across multiple data sets, it may not always become apparent that other previously used data sets may be of potential relevance.”






COMMENTS