Shishank Gupta on Unlocking AI Value Beyond Copilot

Jeff Kavanaugh:

I'm Jeff Kavanaugh, here at the Infosys Connect Conference in Los Angeles. Here with Shishank Gupta, Head of Microsoft Practice. Good to see you.

Shishank Gupta:

Likewise, Jeff.

Jeff Kavanaugh:

Interesting, important topic. Obviously, it's in our sessions, unlocking value in AI. And I think that your area is perhaps, not overlooked, but not fully appreciated because of the pervasiveness of all the Microsoft ecosystem. So where do you see AI making the big impact?

Shishank Gupta:

Now, AI has been there for the last couple of years.

Jeff Kavanaugh:

Yeah, Copilot. Everyone's been using it.

Shishank Gupta:

But as you rightly said, Copilot, your focus in the last few years has always been about employee productivity. How can I make things faster for employees, summarize emails, meetings and so on and so forth. I've always believed that that's just the starting point of unlocking productivity or AI value, not the end point. And what we're seeing now, the conversations now, is about how AI can make an impact on the business, which is about how can I take the learnings from productivity for users to really unlock value on the business processes and what the business does in terms of generating revenue, reducing cost, improving regulatory compliances and so on. So that to me is the real unlocking of AI value. What Copilot did, or tools like Copilot did, was getting users more comfortable with using AI technology, but that by itself was not sufficient. So I think now with agentic AI coming in, that is where the true unlock will happen.

Jeff Kavanaugh:

It's interesting you mentioned that. What do you see with agentic AI coming in now as a new form of, an elevated form of AI? What is the non-technical change it's bringing about in these environments?

Shishank Gupta:

So the key thing what agentic AI is trying to do is not just help people do what they were doing before faster and more efficiently. It is helping users and businesses to reimagine the way work was done, the way business processes were done, create new business processes, create new insights that were possibly impossible for humans to do it or would take such a long time that the impact to business it would have if people did it manually was minuscule. So in that sense, what agentic AI is doing is really, as I said, unlocking value in information and data that existed, in using that more efficiently, and using what you already have to create new ways of working, new processes, new business flows, workflows, so that their business can be reimagined.

Jeff Kavanaugh:

You know, when you think about enterprise AI and some of these tools and models, it's sometimes hard to imagine at the individual level, but in the world of Microsoft, you can already imagine we use Excel and PowerPoint and all these tools and Outlook. What is this bottoms-up dimension, this very familiar dimension that a lot of these tools have? How is that going to help maybe accelerate adoption for AI?

Shishank Gupta:

So we've spoken about the value of AI coming in from the context. And I think the important part with Microsoft that we get is all of us as users, in fact most users across the world are on Microsoft technology, right? Word, Excel, Outlook, SharePoint, OneDrive. And that is a lot of user information and context. Tied to it is also the organization context in which it operates because I spoke about SharePoint and access privileges and so on. And with Windows as an operating system, you know access permissions, privileges, roles, and so on and so forth. And Teams being a good collaboration channel over the recent past. Now that context, when you put it in any AI tool, and obviously Microsoft would want to talk about Copilot, but for that matter, any AI tool, that context gives a huge leverage to an organization to now create personalized journeys for users to do their jobs better because you can create persona-specific AI solutions. Somebody working in operations can create tools and workflows that are specific to them. Somebody working on sales and marketing can use CRM data more contextually. Somebody doing research can use research tools on AI on data that they have permission to to do their work better. So that becomes very, very relevant and powerful.

Jeff Kavanaugh:

I was just thinking that although these models were initially put out there so important, really the models will come and go, but your context and your data, that's in the company and Microsoft's pretty close to that.

Shishank Gupta:

Absolutely. I think the models will do things that are standard and static or largely known to users. Maybe that become a utility over time. But the context is going to come from specific organization and user behavior.

Jeff Kavanaugh:

That's your IP as well.

Shishank Gupta:

That's our IP. And I think that's where the Microsoft technologies play a key role because they're so pervasive in the workplace space. I think they really have an edge. Now it's up to them to leverage that more efficiently, but I think they do have an edge in terms of the access and data. Because data is important and critical in AI adoption and scaling AI.

Jeff Kavanaugh:

The fact that Microsoft products and systems are so embedded, do you think it gives them a head start or a leg up in the privacy and cyber concerns?

Shishank Gupta:

It does, in fact, Microsoft ecosystem I think is very well placed because if you look at the products they have, they have products on user productivity, which is Teams and Outlook and so on. They have products in the data space, which is Fabric, they have tools in the space of cybersecurity which is Purview, Defender. They are one of the hyperscalers with Azure. So, they have access to all the information and the workflows and things that happen. So it does give them a heads up against most other organizations who are either in the space of just creating a model or are only hyperscalers, are only having ERP. So they are limited, but Microsoft is not. By the way, I forgot Dynamics, which is their ERP and CRM product as well. So they do have a real head start in terms of the ecosystem in which they play. Now as I said, can they make the most of it is for time to tell, but they do have a head start.

Jeff Kavanaugh:

Taking one step outward, you also have responsibility for what you call workplace services, digital workplace. What does that encompass and how is AI playing a role there?

Shishank Gupta:

So digital workplace services is all about right from the first experience an employee has when he or she joins an organization, whether it is their onboarding process into an organization, all through their training, their learning processes, what accesses they have, the devices that they use, management of those devices, management of all the services that they use, and the entire ticketing process, the experience that comes with it. So all of that is digital workplace services. Now what AI does is really improve that experience multifold because a lot of those things which are human dependent, prone to error, not learning from past can all be eliminated with AI. A lot of these things are self-help today. They learn from context of the employee, of the organization. They have real-time access to changes in policies and permissions. So it is far more different from what it used to be before. And we see that for our clients. We are helping our clients to reimagine their workflows when users are onboarding. The learning journeys are much more personalized and customized basis their preferences, their patterns, their role personas and so on. When they raise a ticket, most times tickets are auto raised for them or they are able to self-heal and solve the problems themselves before having to go to an agent.

Jeff Kavanaugh:

Looking in this part of the business, what do you see is the most important item nine to twelve months out that if you can get it right will really increase AI adoption or value?

Shishank Gupta:

If you look at it from an organization perspective, the most obvious thing that comes to mind is reduction in cost. However, I think it's not just the cost element. Cost will get reduced with automation and agents. I think the consistency of experience, making sure that the latest information is used all the time, the learning is much more broader and based on data and not individuals who are servicing these things, those are going to be very significant changes that will happen. And that will have a very positive impact on the experience of the users. And when the users experience the positive impact of AI, they are more likely to use it, AI, for their own work because they see it works.

Jeff Kavanaugh:

Kind of a last question, what do you see the role of humans as this progresses and their relationship with these agents and robots?

Shishank Gupta:

I think today the… if you think of the junior roles that we had and people talk about roles will go away, I think they will get redefined. They will all have buddies and agents to work with. So now they become the people who are creating those, guiding them, maintaining them, making sure they don't go rogue and making those final adjustments as they go along. I also believe the roles in the organization will now become more, as I said, mentors and guides to these agents and teams rather than just doers of what they were doing before. So it will be different. There will still be manual effort, but progressively that will go down and most people will start to feel more comfortable working with agents as their buddies rather than looking at it as a competition right now.

Jeff Kavanaugh:

And because they're already familiar with the tools, maybe that will increase adoption and they'll feel better about it as well.

Shishank Gupta:

Absolutely.

Jeff Kavanaugh:

All right. Shishank, thank you so much for your time.

Shishank Gupta:

Thank you very much.

Jeff Kavanaugh:

It's Jeff Kavanaugh. Until next time, keep learning and keep sharing.