Heard at Ops: Creating Strategies for Data with Greta Lovenheim

As part of the Heard at Ops series, Greta Lovenheim of PwC describes the shifting skill set for operations professionals as well as the potential and considerations of AI from a data perspective. Watch the video and read the Q&A below.

Q. Describe the shifting skill set for operations professionals.

A global survey was done about the key skill sets that firms are looking for – that they see as being needed – and data science, across most territories, was number one. I agree with that.

But as firms go and seek that skill set, in addition to the more softer skills and business knowledge, what I’ve seen in terms of a pitfall is a firm will find some good resources and then the queue forms asking, ‘Can you help me get this data?’. That takes over the data scientist’s day. So, it is important to hire, train and cultivate, but also to manage data science resources: to make sure the data requests are not taking over what else they are supposed to be doing or to be sure that the request is answering the questions that need to be answered.

Q. What is the potential of AI and its impact on data?

There is a lot of buzz about AI and generative AI and how it impacts the future landscape. One of the things that I’m excited about potentially for where generative AI could go is it might democratize the ability for people to get into the data. Today, if you know how to code, that’s your toolset; that’s how you get in and help get answers to your questions. But you have to know how to code to do that.

With generative AI, potentially – and I don’t think it’s that far away – you may not need to know how to code. You still need to know how to process the questions, how to think about the answers, and understand what the data means. But to get access to that data, to start to put that puzzle together, you may not need the code going forward to be able to do that.

Q. What should operations teams keep in mind about AI?

AI means a lot of things. What came up [in the panel discussion] and what is certainly a factor with AI is the explainability. In terms of pre-Chat GPT, if you’re looking at decision trees or neural networks or cluster analysis, it is all very useful and it solves very complex problems. But depending on what the problem is, how much of it are you going to need? And how transparent is the process to be able to explain it to your customers, customer-facing employees, regulators, and others? That is a challenge.

With generative AI and what inherent biases may exist, that’s where I think there is a long way to go. It’s moving fast, so it is important to understand and know the biases and to focus on being able to explain what is happening. That is a huge realm ahead.

Greta Lovenheim is a Partner with PwC’s Customer Transformation practice focused on Customer Data and Analytics.