Main content

Qualtrics AI Chief - ‘AI is the new UX and it is frictionless’

Derek du Preez Profile picture for user ddpreez May 2, 2024
Just a few days into the job, Gurdeep Singh Pall joins Qualtrics from Microsoft, where he is now heading up the experience management vendor’s AI strategy.

Bot with speech bubble, concept of chatting with bot. Asking question to AI. Question mark and bot representation ready to work © SaskiaAcht - Shutterstock
(© SaskiaAcht - Shutterstock)

Gurdeep Singh Pall was appointed as Qualtrics’ President of AI Strategy just a few days before the experience management vendor’s X4 Summit in Salt Lake City this week. He joins the company from Microsoft, where he had over 30 years of experience, and was most recently part of the Technology and Research leadership team that helped spearhead Microsoft’s OpenAI partnership. Simply put, this is someone that has been very close to the latest developments in generative AI and has a thoughtful understanding of their impact on the enterprise. 

So, why did Pall see Qualtrics as his next career move? As someone that has had front row seats to the latest AI developments at Microsoft, a company that in many ways is leading the charge, his options wouldn’t have been limited. Pall said: 

Qualtrics’ ten billion [experience] records - I mean, this is gold. This is one of the reasons why I came here - the assets that the company has are pretty valuable. 

This week we’ve seen how Qualtrics plans to put this collection of experience data to use, with the launch of its AI tools to support its XM platform. Once a vendor that focused on surveys to understand employee or customer sentiment, Qualtrics is now pushing heavily to become a system of action, one that can change how an organization interacts with people in real time, based on how they’re feeling. 

The concept of ‘experience’ has played a critical role in enterprise technology decision making over the past decade. diginomica has written extensively on the topic of ‘frictionless enterprise’, which essentially is about improving the user experience to enable a variety of improved outcomes. 

This begs the question, what role does AI play in ‘experience’ and further reducing friction? According to Pall, generative AI will fundamentally change how we interact with technology and organizations, where AI is the new UX. And that UX is adaptive, personalized and ‘has no friction’. 


Pall said that, unsurprisingly, the next decade will be defined by AI and that Qualtrics is a “unique place” to focus on the advancements in AI models. Providing some context, he said:  

What a lot of people don’t realize is that the only example of a ‘pure AI app’ that existed before this decade was search. If you look at all the AI algorithms that Google is using, or that Microsoft is using with Bing, they are really about learning peoples’ preferences and building these models. Their advertising models are so good, because they know that the wrong ad means less money. They have built incredible amounts of intelligence. 

However, he argued that these models are locked in a ‘search silo’ and that nobody has the ability to take those advertising models and apply them to other things. This is particularly true of enterprise functions, such as customer support, sales or retail. However, Pall believes that Qualtrics’ XM platform could play a critical role in this. He said: 

Qualtrics can be a horizontal platform that brings knowledge about people and their preferences, and what they want to do, and apply that to lots of industries. 

You can look at Qualtrics a little bit like the engine that existed inside Google and Microsoft, and even Facebook, where they were able to glean a lot of ‘signal’ about users and then use that for some purpose. In their case it was advertisements. 

But here, you can use this platform for driving the direction of customer care, a customer contact center conversation. Or if it’s an upsell situation, how you can drive it towards that. 

A world of empathetic agents

So, what does this world of AI as the new UX actually look like? Pall believes that we are embarking on a fundamental shift in how we think about engaging with organizations and the experiences we’ve got accustomed to on the Internet and through mobile technologies. Critically, he believes that the assumption has always been that humans are smarter than machines, so experiences were designed using workflows or pathways to guide humans to what they need in the most convenient way possible. However, if we are entering an era where machines are as smart (if not smarter) than humans, then that changes how we think about designing these experiences. 

As noted, Pall believes that AI will take us to a place of “no friction”. He explained: 

It’s a different medium - you’re not fighting with friction and trying to make it less. There is actually no friction. Whether it be web or mobile, we created experiences that served a limited number of intents. When you come into the Delta app, for example, and you want to book a flight, or change a flight, or look at points - you can literally write down the number of ‘intents’ for why a user is coming to your app. 

Your goal is: how do I quickly identify what they want to do so that they can guide themselves through the steps? That becomes the multi-step, task-based completion of what they want to do. You can see the difference between the ‘good apps’ and the ‘bad apps’.

This will change as generative AI becomes more sophisticated. He argued: 

“What if in the AI world, you fire up the app, and you don’t see anything, other than what we believe you are here to do? At this point I’m in the predictive world, so I better get it right nine out of ten times, or 99 out of 100 times. 

But you don’t see any of the clutter. There is no ‘my wallet’ or ‘know your trip’ or anything like that - there’s nothing. You just come in and they know what you’re calling for, because they have enough ‘signal’ that will let you zero into that experience. 

After that, you come in and express what you want to do, in whatever way, and the entire UI is completely adapting itself to the task you want to do. 

It’s quite a compelling pitch, although no doubt some will be skeptical that organizations will be able to understand their customers or employees well enough to tailor experiences in that way. And of course, there are a number of justified concerns about data visibility, biometric data collection, and processing of data in this way. However, this is new territory and it’s undeniable the direction of travel that we are heading in. 

Pall envisages a world where instead of the traditional Delta app, a user would be able to converse with a ‘frictionless app’ and say ‘I’m traveling to Morocco’ between two dates, and then be presented with flights, hotels, transfers, restaurant recommendations - all based on your particular, individual needs. He said: 

With AI, in whatever way you can express your intent, I can tailor the experience exactly for that. It’s predictive, completely personalized and adaptive. 

Everyone’s focused on Large Language Models or the latest GPU from NVIDIA, but the more interesting thing to me is how UX is going to completely change. 

If it’s an agent, it should be working all the time for me. In a much deeper way. It’s not transactional, it’s completely adaptive, and it knows me in a highly personalized way. It should even give you explanations of why things go wrong, like a knowledgeable employee would. That AI is ready today. 

The enterprise impact

It’s a very interesting concept. It feels far fetched, and yet the pace of change we are seeing in the industry today has been staggering. I don’t believe this is something that we will see in the next year or two, but it certainly feels like this is what buyers are aiming for. Pall argued that these AI agents will be so individual, that they will cater to each of our defined needs and wants. He went as far to say that, given that the drivers behind the models will be different, it will feel like these agents are empathetic. Pall explained: 

I believe that the AI experiences that we will see as the next big wave of post-mobile apps are going to be these very empathetic agents. The machine is built in a way that is designed to drive certain things: customer satisfaction, loyalty, long term value - these are the metrics that the models are trained on. They will get highly empathetic. 

It’s not hard to see why Pall believes Qualtrics could play a critical role here. These drivers - customer satisfaction, loyalty - fall squarely into the vendor’s wheelhouse and have been the focus of its AI announcements this week. But this begs the question, one which I’ve been pondering in recent weeks, what will the impact be on existing enterprise technology systems and how enterprises organize themselves? I’ve written on this previously, but it’s fairly logical that if users aren’t interacting with systems in a structured, process driven way - one that’s often defined by the existing structures of an organization - does that mean that we are going to have to rethink how enterprises operate? Pall thinks this may be true. He said: 

I think your point about organizational structures is actually quite profound. I’ve been thinking about it. Let’s look at two particular areas, say marketing and customer support. There is no reason why those two should be separate. There is absolutely no reason. I think that is what will come out of this, what I call ‘AI-based experiences’. 

And new AI models will have a deep impact on the existing enterprise technology stack. Pall added: 

What is the role of OpenAI in the tech ecosystem? I think they are going to be an accelerator for the next generation of enterprise apps. But you have to come in with the right mindset.

If you have lots of silos and complexity, and you think then that you’re going to put in this little AI engine, and everything's going to work fast - that’s not going to happen. It’s like saying I'm gonna put a turbo engine on my bullock cart. It's not gonna go anywhere, right? 

Because what happens is, if AI needs to interact with your structured data, where one query takes a minute for it to come back, you're not going to do much. So you need to rethink your stack.

My take

This was a really enjoyable and thoughtful conversation. Pall has extensive experience in this area and you can see he’s clearly thinking a few steps ahead. The reality for most enterprises is that they are quite a way off this future - Pall’s comments about AI working with existing structures and technology stacks is spot on. A lot of work is going to need to be done to make that data accessible, workable and for organizations to manage the change towards AI enabled experiences. I also have personal concerns about how data used to deliver these AI experiences is collected, accessed and processed (for Qualtrics’ part, it says that its customers won’t experience any ‘leakage’ to public AI models). But there’s still a big difference between data visibility in a form or workflow and data visibility processed in a model. We need to think that stuff through. However, I can’t deny that the world that Pall is describing feels more tangible than it ever has and we are going to see some interesting changes over the coming years. 

A grey colored placeholder image