Bringing AI to the Enterprise
- By Paul Mah
- June 06, 2023
Slowly but surely, AI is making its way into the enterprise. Business leaders are rightly cautious, given its novelty and complexity. Beyond this veneer of caution, what is really stopping businesses from fully utilizing AI to its full potential for crucial business decisions and driving growth?
At a panel discussion at the recently concluded 6th Chief Digital and Data Officer Summit, executives well-versed in the realm of data science and analytics shed light on this topic. Drawing from their experiences in this red-hot niche, they discussed what can and cannot work with AI, how to successfully integrate AI with automation, and the surprising internal pressures they face from AI teams.
Bridging AI and automation
While AI and automation are two technologies that can help businesses move the needle if rolled out as separate initiatives, they shine the most when deployed together, says Dr. Jingyuan Zhao, the Group Chief Data Officer at Great Eastern. “If we can combine these two, I think the benefits can more than double… we will see scalable business impact.”
Dr. Zhao used the loan approval process for a new customer in the financial services sector as an example. Instead of rolling out AI as a standalone solution to approve loan applications, why not integrate it with automation to amplify both the customer experience and employee productivity?
“The chatbot can collect the relevant information from the customer, which is then sent to the AI engine for approval. The result is then seamlessly fed back to the customer, delivering an exceptional experience to customers and productivity for employees,” she said.
“The intent behind automation is to get things done faster. AI adds the cognitive layer to augment enterprise intelligence for doing things smarter. And organizations that can do things faster, and smarter are going to win,” noted Deep Thomas, Group Chief Data Officer at Nomura.
“[Using] both of these together can give you a multiplier effect. They can still be effective individually, but automation drives efficiency, whereas AI drives effectiveness, and you need to have both of them together to get the multiplier.”
One more tool in the toolkit
As with any new technology, the dangers of hype are real. For its incredible capabilities, AI is no silver bullet. For Nikolay Novozhilov, the head of Data Science and Advanced Analytics at the Bank of Singapore, the latest AI models might be more powerful but are also limited in various ways.
“I look at them more as building blocks, there are certain things they can do and there are certain things they can’t do… I treat it as [another capability in my toolkit]. We are certainly working on developing AI tools to bring value to the business. But what businesses need to figure out is what is usable and the business processes that AI can impact the most.”
“A lot of times, AI is compounded with digitalization. You have to be clear on the right use of this technology. Only then can you create a strategy around it,” said Dr. Martin Saerbeck, the chief technology officer of Digital Service at TÜV SÜD.
“What kind of people do you have on your team, what are the milestones you want to achieve? It's not just demonstrating that it works, but integrating it into your business. And just like every technology, you have to do your risk analysis, risk profiling, and mitigating potential risks, continuously."
The pressure from within
Surprisingly, the greatest pressure might not come from the board or other senior executives, but from team members raring to go.
“I feel there is quite a bit of pressure from the data scientists themselves; they want to play with AI, right? So there is this pressure from them, but you have to start with the business value. If you find a very valuable business process then AI can come in later – but there is certainly this pressure from that from engineering,” said Novozhilov.
Thomas agreed, noting that the priorities of senior executives typically revolve around cost avoidance, operational efficiency, risk mitigation, enhancing customer experience, and generating more revenue – among others. "Whether you use it through deep learning models, or you do through traditional analytics is not a discussion area.”
“It is our responsibility to not get enamored by the advanced concepts that AI has to offer, but to stay focused on the [business] objective. Sometimes the biggest impact can be generated through a dashboard with some level of insights and alerts or a basic segmentation or a simple decision tree. As leaders, we have to also effectively harness the aspirations of our data scientists and data engineers that address business needs and generate value as opposed to only pursuing things that may deem disruptive or futuristic that perhaps the Big Techs of the world do,” Thomas explained.
But even organizations that face pressure from the management to implement AI initiatives should take it in stride. This is a good thing, says Dr Zhao.
“The top-down strategy is very important for any new technology. If the top management don’t adopt new technologies and doesn't make any investments, then it would be a very difficult journey,” she summed up.
Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at [email protected].
Image credit: iStockphoto/Andrii Yalanskyi
Paul Mah
Paul Mah is the editor of DSAITrends, where he report on the latest developments in data science and AI. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose.