Operators are in various stages of implementing artificial intelligence (AI) in their businesses. As projects proliferate, they will need to establish a common AI strategy for their entire organization. But just knowing how to start can be daunting. AI experts from global operators Axiata and Deutsche Telekom and business software and IT services company Etiya shared their experiences and lessons learned from deploying AI at scale during a panel discussion at Digital Transformation World Series 2021.The key to getting your AI strategy right is to “not be scared to face the fact that it might not be right,” said Angela Maragkopoulou, Senior Vice President B2B and Data Analytics at Deutsche Telekom IT GmbH.
Deutsche Telekom is working on embedding AI and advanced data analytics in multiple practices across its business. Maragkopoulou said the first steps were to get people involved and “passionate about use cases around AI” and establishing a framework within the organization to empower data scientists within the businesses to create more AI use cases.
“We created a central function that checks the strategy as well as the capability and the skill sets … That’s been quite successful for us,” she said.
Axiata, the multinational operator group based in Malaysia, has a similar AI strategy to Deutsche Telekom that it has developed over the last five years. The process starts with identifying the use case, then identifying where are the data sources, and whether real-time data is available, explained Dr. Keeratpal Singh, Chief Data Scientist at Axiata.
Singh said it was important to have a “top-down approach” because sometimes the teams of data scientists have challenges with getting the right data. For example, sometimes real-time data is not available, which could be essential to some use cases.
Both Axiata and DT said they had also established academies for training analytics professionals and data scientists to ensure that they have the skills they need, noting the importance of reskilling and upskilling their own employees as well as attracting external talent.
Apostolos Kallis, Chief Commercial Officer at Etiya, said one of the challenges for scaling AI is that operators are often too focused on the return on investment (ROI) and want to be able to prove how AI will either improve customer experience or their bottom line.
“Service providers need to understand that this is a long-term commitment. You need to start small and be open to the fact that the benefits will not manifest themselves immediately,” said Kallis.
He also observed that operators need to be more open to failure when it comes to implementing AI at scale, but noted that they’re not used to thinking like that. Another important consideration is winning executive level support and commitment across the organization.
“It can’t be one single person’s or one department’s job to implement AI”, he said. “I think operators are realizing that this is a company-wide initiative. So there need to be KPIs in place, incentives in place, to make sure that this is on everyone’s agenda and the whole organization is moving in that direction.”
Maragkopoulou said that getting everyone on board with AI “still needs a lot of effort” and “education” because it is a technology where there is a “very big gap in understanding,” not only in what AI does but in how it will affect the business.
The starting point for making the business case for AI is a sound data analytics strategy, said Maragkopoulou. For the last three years ago, Deutsche Telekom has been “un-siloing” and democratizing data to make it accessible across the organization. Once the data strategy is established, then the next step is to ensure the right skill sets are in place and focus on “big bet” areas of AI that could measurably deliver the biggest ROI.
“After that you have a very good baseline and understanding to start convincing [others] how we can invest more into this technology,” she said.
Watch the panel session AI at scale - Ensuring an organizational wide approach here.