It is an era of digitization and investment. Companies are investing in AI technology to transform their businesses.
However, despite digitization efforts and investments in AI technology, businesses are still facing new challenges nearly three years after the start of the coronavirus pandemic. From shifts in consumer buying habits to employee turnover to supply chain issues, companies are looking for ways to address these challenges while remaining relevant in the age of digital transformation.
Companies are not sufficiently prepared to meet these challenges, said Sateesh Seetharamiah, CEO of Infosys subsidiary EdgeVerve, a provider of robotic process automation technology. He is also a member of the MIT Auto-ID Lab, an IoT research group.
In this Q&A, Seetharamiah explains what prevents companies from effectively applying AI technology to their business problems.
What is one of the biggest challenges companies face when implementing AI technology on the path to digitization?
The real deal with AI and all the value it can create is having contextual information and data.
Sateesh SeetharamiahCEO, Edge Verve
Sateesh Seetharamiah: While everyone is interested in AI, I think the real problem with AI and all the value it can create is having contextual information and data. And without that, AI’s ability to influence cognitive operations and decision-making will be limited.
For example, underwriting in the insurance industry. Underwriters must accurately assess risk, approve or deny insurance claims, and approve the insurance itself.
The amount of information they need to make this decision is enormous: enormous monetary, historical and claim information. Many of them are in digital contracts; many have paper contracts and video files. The time they spend on decision-making is a fraction of the time these people spend collecting the data and organizing the information they need.
There are many opportunities to use technology to digitize information from documents, bring it in and create the data layer.
Second, if you have that kind of contract data, the AI that goes with it can make relatively smarter recommendations to those underwriters, in terms of what kind of risk is acceptable. But AI today is hardly used in this process as they do not have relevant data that can be fed into the AI technology to make relevant decisions. So a very central insurance process…is completely manual in human decision-making.
There are many cases where we find that unless and until everything is digitized, unless and until all of this data is very contextual to this business…it’s very difficult to apply really AI at a very operational level.
What kind of processes are needed to apply AI in the scanning process?
Seetharamiah: There are so many policies that govern every process in a company. Not only must AI have access to the underlying data, but it must also have access to key decision-making policies – and how policy has historically been interpreted by humans and applied to their own decision-making. .
The first is the policy, and the second is the interpretation of the policy and how that translated into the decision-making. This needs to come together if AI is really to make recommendations that businesses can actually use.
Most companies don’t have the know-how to know how processes are performed, let alone automated. Experts in these companies have an idea, but do not know how these processes are carried out in several companies. Many of them are done even in shadow IT – it’s not even part of the main business landscape. So that led us to say that we need technology to decipher how processes are executed.
We need to apply technology intelligently to digitize as much as possible. We need to ensure that digital information can be translated so businesses can ultimately benefit from it. It must serve a larger commercial cause.
What’s stopping companies from doing this kind of AI digitization?
Seetharamiah: I think the challenge is really having the right kind of governance mechanism between the business and the technology, because ultimately that’s where the real ideas and the real implementations and the translation of technology into business benefits.
It’s not just a matter of technology. It has a lot to do with internal governance, internal structuring and the right mindset in approaching this technology. I don’t think any company wants to implement even a single dollar of technology for technology’s sake in the end. They must all show a lot of business needs.
Obviously, there are apprehensions about some of the things AI can exploit, such as “Can I hand over this decision making?” Will this affect my work? There are so many other concerns there.
There’s a bigger problem to solve, which goes beyond technology, which has to do with mindset, which has to do with adopting this next-gen technology and convincing people how it can not only influence the business, but can also influence theirs. individual performance.
Editor’s note:This interview has been edited for brevity and clarity.
