The Challenges of Integrating AI Into Your Business | Salesforce CIO Corner

There is a resistance sometimes to change. And part of that is because people don't want to be commoditized. They don't want their hard earned expertise to be extracted and put into a system and then mechanized. I think part of it too is that people wonder, well, if there's a system doing the things that I do, then what do I do?

Welcome to the CIO Cordoner. Here's where we will talk to leaders, IT leaders about the biggest challenges and opportunities that we all face when it comes to integrating technology in the enterprise. And today I'm really fortunate to be joined by Casey Coleman. Casey is an SVP here at Salesforce working in our public sector practice, but also Casey was the CIO of the U.S. general Services Administration. One of the things that I wanted to start the conversation with was understanding what you're hearing from CIOs that you currently work in, in your position here at Salesforce about all the challenges of building a cohesive data and AI strategy.

Yeah, Juan, I know you talk to CIOs all the time, so do I. Everyone is excited. The potential in the public sector especially is enormous. We can envision AI solving really complex problems that we just couldn't address before, like much more rapid introduction of life saving drugs or smart self healing, traffic management systems, improvements in energy, education, economic development and more. But in every single case, the key issue is data. All of the AI systems that we're using now have been trained on publicly available data. And in every case, the government use cases need government data. And that sits in a lot of different systems in every format imaginable from decades ago. And so getting to that data, using it safely and securely and with an ethical framework so that we're protecting life and health and safety, these are the challenges that government CIOs are addressing. And so they've got to proceed with with a lot of care.

You know, when we think about data over the years, organizations, whether you're in the public sector or not, organizations have been building all these different silos when it comes to data that now have trapped this data in ways in which, you know, in many cases it's difficult to get access to it. But we all collectively know that if we want to get the maximum benefit from generative AI, having all the data and in a state that can be used is critical to success.

What is it that you're seeing in the private sector and with all the CIOs you talk to about how they're building a successful strategy with their AI.

And data first of all, it's always understanding and assessing the current state of your processes, understanding where you have ultimately opportunities to implement AI to the point that you can get significant business value. The second one is to investigate your data. You gave us a really good view as to how critical it is to investigate your data and understand that your data is in the proper state to ultimately get good results from what you see in the AI models that you apply across your enterprise. Important to consider scalability and security. One of the things that I have seen and many CIOs have expressed to me is that they've been working on all these different pilots. Some of the pilots have been successful, others haven't, but those that ultimately were successful, now they're finding challenges with being able to scale and of course secure those pilots for general use across the enterprise. Now one of the things that I have noticed as we've gone fast towards the implementation of AI is that in many cases we've gone out and deployed the technology without having the typical steps that we take in implementing technology at the forefront. And one of those is definitely making sure that we have an effective change management plan. So that's what I'm hearing, that's what I'm seeing. But now you have some really unique experiences, Kasey, that I think are important for our audience to hear about. And specifically in the area of general services, you managed as the CIO, a group of more than 12,000 employees nationwide. Give me a bit of a sense as to some of the specific areas where you think this type of technology, specifically generative AI, can have a significant impact.

People go into the public sector to make a difference and to give back and to do important work that serves, serves all people. And when they get there, they often find that they're stepping back in time with a set of technologies that are from way in the past. And so as I think about my own experience as a cio, I would really look at the day to day work of our employees and where are they spending their time and how do we take friction out? One of the things that everyone in government does is purchase or procure things. And that's a complicated process. It's really highly regulated and it takes a long time. But it's also the kind of thing that's very standardized. So applying AI to that, you can speed it up and you can give people time back to do the high value work that only human experts can do. So that's where I would start is with those common back office procedures that everyone has to do. And that are pretty standardized.

You made me think of an example here at Salesforce that I use often in my discussions with other CIOs. Simple problem. You're an employee in the company. You have to get an answer to a question. It could be a question related to the business, it could be a question related to benefits. It could be a question related to your technology, the computer that you're using to do work in the company. And to get an answer for any of those questions in the past, the employee had to go to multiple places and search for the answer, spending countless hours, in many cases, trying to get to the right person or the right answer specific to a question they had. What we've done recently using Genai is we actually took data from HR data from IT business process data. We brought it all together, and now we have a mechanism using our Einstein AI engines to be able to provide answers to employees using natural language and getting responses much more effectively and rapidly. That has actually reduced so much friction to the point that you made a minute ago in terms of how people do work.

I've got to jump in on that, because I have. I have used those new systems. I think it's called Basecamp. And yes, I. I just got a new phone, a new work phone, and I had to activate it. And so I ran into a couple of hiccups, and I just went to Basecamp and boom, I was able to solve it myself. And this was late at night, so I was in a hurry to get it solved and wrap up for the day. So compare that to my experience as a CIO in government, where all of your systems are in different places. And even the simple act of creating an employee portal to bring all those systems together and applying AI to solve them faster, that's an amazing use case that would just supercharge government work.

Well, you've spoken about the notion of cultural resistance to AI. What does that mean? If we find that in our places of business where we do work, we find that type of cultural resistance to AI, how do we fight against it?

So you've got to find a way to understand and empathize with that and bring people along. But I look at the work that we had in front of us when I was a cio, and it was like trying to dig into a mountain of work with a teaspoon. There was so much more that we could never get to. And I really think if we can free people up from a lot of that manual process work, we can get to those critical systems, like being able to help families and children much more effectively, being able to provide support services for the most needy. All of these are things that really motivate people. At the end of the day, those are the things that public service is all about. So what about you? What's your take on that?

Absolutely. If you went back to the example that I mentioned and that you actually reinforced for me, think about it. You know, if you have not been in the enterprise for too long and if you have not read any of the communications related to that particular solution when we deployed it, you would never know that Gen AI is running in the background. You would just simply know that, hey, that's how we do business here at Salesforce. You go to basecamp, you get to ask your questions, you get the answers that you need, and that's it. That's all. Reality is that in the back we have some really good tools that are using Genai to be able to provide better answers to our employees. But I think that that's what's going to happen over time. We're all going to be using technology solutions that will have AI embedded in them. People will not even know they're just simply doing work. And that in a way, to your point, that will help us deal with any cultural resistance to AI that we may see. Because it's just simply the way that we do work. And by the way, as that happens and becomes more pervasive in the industry, I think we are going to see improvements in productivity that will happen just naturally without having to be constantly talking about generative AI.

You know, you make a great point, Juan. That I think is critical to our own internal success with AI so far. And that is putting it in the flow of work. It can't be just another screen to go cut and paste data. It's got to be part of the way we get our work done because then it's organic and then it's easy and people will want to use it. And I'm sure you see that with the customers you're talking with as well.

Absolutely. Now, you gave me a great opening to give you my perspective on implementing AI with technology that it's already AI compatible, let's call it that, and that it's being used to run your business processes versus having to create and build AI type of solutions that are not embedded into the technology that you use to support your processes. That's much more complex. The fact that we are now embedding AI into sales cloud, into service cloud, into marketing cloud, and now when people who are managing their sales operations or who are actually providing service using our technology, they will have AI embedded in their technology, making it much easier for for that AI to provide value in the business processes that our customers are executing every single day.

And I'm excited about what we can do with AI and data because I think we have an approach that with data cloud and with Einstein that lets us help customers get to all of that data and not rely just on large language models, but be able to use the LLMs in conjunction with their private data that's safe and secure and grounded in ethics.

I agree. Now, along those same lines, from your perspective, Casey, why do you think it's important for us to actually think about AI as purpose built versus using consumer first AI in the enterprise?

Well, think about the complexity and the uniqueness of some government use cases. For example, a public defender and that attorney who is representing that client probably gets word of that case with a day or a few hours in advance before they have to show up in court. So they can't. They need a summary of what case the case is about and what other legal actions their client has been involved in. But they can't turn to a large language model and ask that question. It doesn't have that data. And you don't want that data exposed to an LLM. So you need a way to be able to use AI in conjunction with, in this case, all of those legal files from your internal case management systems. And when you do that, you can bring the best of both worlds. And that attorney can see at a glance a summary of all the key issues. They can understand how to bring a good defense and they can start to bring better representation for the people who are most in need. That's a real use case that I'm hearing about from some of our clients that I'm super excited about. And I'm hearing more in other similar places like with child support, child welfare, family services, and much, much more.

Can you imagine how much benefit using these types of technologies will be generated for so many people in the community that need more access to these services faster, with less wait time, with certainly better quality service at the end of the day? Now you made me think of something that I think is really important as well. It's important for the CIOs to also be what I think of as chief education officers when it comes to Gen AI. We got to be really, really active in educating people as to where the benefits are, what the technology is all about, what types of problems the technology can help us solve, and make sure that both at the management level and of course, even at the board level, there's a good understanding as to how this technology works and what it does and what it can and cannot do. At the end of the day, whatever we do when it comes to technology, it needs to be closely aligned with the business strategy. So to me, nothing changes here. We still need to have a solid business strategy to drive growth, to drive efficiency, to drive better products and better quality for our customers, to improve our customer experiences, to improve our employee experiences. So that doesn't change. Now we have, of course, our technology strategy that follows. And what's happened now is we have another great tool called Generative AI. And what I think we need to do as CIOs is embed those generative AI capabilities into our technology strategy so that we can support the business strategy as a whole. And if we continue to do what we've done so well for so many years in terms of creating and driving business value through the use of technology, now applying another tool in our bucket, I think we'll be just fine. But for that, I truly believe that we all collectively need to understand what Gen AI is capable of, where are the best places to use it, the appropriate, ethical, secure way of using Gen AI, and like you said at the start, making sure that we have the right data to support it all.

Thanks, Juan. Awesome conversation. I appreciate it.

Well, Casey, thank you so much for joining me today. And please check out the links in the description below and be sure to stay tuned for more episodes of the CIO Corner. They will be coming soon.