Charles Lamanna on AI’s Next Big Role

MOLLY WOOD: In just one year, Microsoft Copilot has changed the way we work forever. By now, business leaders understand how it can boost their individual productivity and the efficiency of their teams. But as generative AI evolves, a bigger and more consequential opportunity presents itself: total business reinvention. Yeah, buckle up. In today’s episode, Charles Lamanna, Corporate Vice President of Business Apps and Platforms at Microsoft, goes beyond what’s possible today and shares what the near future of AI looks like. We talk about low- and no-code tools, and how AI is evolving from being just a personal assistant to being a group assistant. And of course, what business leaders can do to prepare for these exciting new capabilities. Charles has led an incredible career. He joined Microsoft right out of college as a software engineer, then started his own cloud monitoring company, MetricsHub, which was then acquired by Microsoft. He rejoined in 2013 and has since led the charge on some of Microsoft’s most exciting new products. Here’s my conversation with Charles. 

[Music]  

MOLLY WOOD: Charles, thanks so much for being here with me.  

CHARLES LAMANNA: Of course. Thank you for having me.  

MOLLY WOOD: Let me start by asking you about the portfolio of products you’re working on now, because you have been at the center of what will be two massive transitions, from local data to cloud and now pre-AI to AI.  

CHARLES LAMANNA: Like you mentioned, there is a big transformation for business applications, a business process where you went from mainframe to client server architecture, or from client server architecture to cloud. And that was all very much about the hosting and IT management aspects of business apps, not as much as how the processes themselves were run. I mean, the same way you record a procurement or a payment, it’s been the same for 40, 50 years. AI, though, we think is going to fundamentally change that because it’s not going to be the same type of apps and workflows just moved to a new hosting environment. But instead, it’s going to be fundamentally different workflows. And we kind of have this vision of people and copilots working together to complete tasks. And instead of a highly repetitive, structured, predefined workflow to moving to a world of highly dynamic, highly reactive, highly agile workflow and processes, with people being augmented by copilots to really be more productive than we’ve ever seen before when it comes to business process and business applications. 

MOLLY WOOD: I’m going to ask you a million more questions about the specifics of that as one of the few people who is really, you know, on the inside and sees what’s coming. But before that, can we dig a little deeper into the idea of low code and no code? Because I think this is—I was at a party recently where somebody said, ‘I’ve been trying to teach myself Python, thinking I’m going to need it to interact with LLMs and AI, but maybe I don’t.’ 

CHARLES LAMANNA: Yeah, absolutely. So my background’s as a developer, so I love writing code, but I recognize there’s seven, eight billion people on Earth, and there’s like 30 million people who write code with regularity. And what’s kind of unfortunate is so many great ideas exist out there to improve people’s lives, improve business process, and improve, kind of, just the world, but they’re bottlenecked by people who can write code. So what low code or no code is all about is this idea of, what if instead of making people learn how to program, what if we made programming accessible to everybody? And we talk about this idea of Clicks Not Code. So you can drag and drop and build solutions visually, or if you want, you can go drop into lightweight expressions as opposed to having to use fully fledged code. The analogy I always make is it’s like PowerPoint and Excel had a baby. It’s kind of what low code is all about. This has, as a result, contributed to rather substantial large-scale adoption of these low-code tools inside the enterprise, inside the workplace, where people can now build apps and workflows and visualizations and reports that they need to get their job done and don’t get stuck waiting for a coder to have the time or for them to find the budget to go build the solution. And this idea of democratizing technology is what computing has been all about, all the way back to the mainframe, to the personal computer, to the smartphone. This constant trend of things becoming more accessible and requiring less skilling and training to use software.  

MOLLY WOOD: Could you describe one? Could you give me an example of, you know, something that you could build with Clicks Not Code that you found particularly powerful?  

CHARLES LAMANNA: One of my favorite examples is a guy by the name of Samit Saini, who worked at Heathrow Airport. He worked in the security group, so he would, you know, help run the policies at security checkpoints to take your liquids out of the bag, or take your belt off to go through the scanner, that type of thing. And no programming background at all. He was able to teach himself low-code platform Power Apps using videos, and then he was able to build a bunch of Power Apps to remove paper from the security process, because he was very motivated to get rid of these big thick binders that would be two, three inches thick with tons of different translations, because you have to have all the different languages when people go through security, or all the different protocols and processes, and he thought there has to be a better way. This should be on my phone, not in a binder. So he learned Power Apps, he built a Power App, and that’s what the airport was able to ultimately use to digitize. I love this story for three reasons. The first, the goal is great, it’s righteous. Eliminate paper. That’s better, you know, just for so many reasons. Number two, Sumit was able to elevate his career. So he now works in IT doing full-time power platform development, even though he didn’t study computer science. And if you asked him a few years ago, what’s Python, he’d think of the animal, not the programming language. And then the third bit is just this idea that the airport itself runs more efficiently. So, it’s rare. You’re doing good to the environment, you’re doing good for people’s career, you’re doing good for the business. All are winners. And that’s kind of what, at least for me, gets me out of bed every day with excitement and energy to come to work, because you see this ability to, across so many different dimensions, make a difference through technology. 

MOLLY WOOD: Right. Yeah. I mean, it makes me wonder what I could build with Power Apps and low-code tools. I mean, speaking of accomplishing more by doing less, it seems like the data backs that up, right? The 90 minutes of time savings per week for sellers who are using Copilot. A 12 percent increase in overall customer satisfaction. Clearly, you’re a big thinker. Tell me what else you see in the AI transition. You know, walk me through what you think is going to be possible that maybe people who are just experimenting with this aren’t even seeing yet.   

CHARLES LAMANNA: One of the things that gets me really excited is the creation of new types of jobs that require business expertise but start to have, kind of, career opportunity and scalability like a programmer does historically. One of the things we’ve seen around Copilot and customer service settings, one of the most important things to a successful rollout, is having curated, high-quality content. Because Copilot reasons over all of your knowledge base, your help articles, your onboarding docs. And it does a great job reasoning over those and giving a really precise answer for users. But if the content that it has access to is old, it’s stale, then Copilot is going to give you stale answers. So what we’re seeing is there’s almost this content ops role starting to appear, where companies are creating dedicated teams whose job is to curate, prune, and improve the content that feeds into Copilot. The job is to build the right content that will make Copilot work great, but you don’t have to know how to write code. The idea of, like, how do you empower more people to contribute to the AI and digital economy? This is a great example of it. So I think we’re all going to have to embrace new roles, new team structures, new ways of working that go beyond just making everybody individually more efficient and more productive. 

MOLLY WOOD: Talk about some of the other, like, the pillars of that transformation, right? Automation, collaboration, customization—what are you seeing in those buckets?  

CHARLES LAMANNA: Historically, Copilot has been really focused on a person privately talking to their AI companion, kind of one on one. But we’re kind of opening the aperture to make it where a single person or multiple people can engage with one or many copilots concurrently. The benefit of this being, you start to have new team composition where Charles and Susie and John are going to work with the sales copilot, the finance copilot, and Microsoft Copilot to get the job done as quickly as possible. If I were to kind of go back to the first one, around automation, this is kind of my personal passion of Copilot this year...  

MOLLY WOOD: Dig in.  

CHARLES LAMANNA: Because, yeah, what we’re seeing is there’s always been this push to automate more of the tasks that people complete every day at work. And there’s just so much monotony and drudgery that people have to sift through. You know, everybody has the job: fill out the time card, copy-paste the data from system one to system two, take this information from a dashboard and then convert it to an email and send it to your boss every Friday afternoon. Those things are not what we should be spending human creativity and ingenuity on. That’s a great place where Copilot can start to automate those tasks. So, what we’re announcing is this idea where Copilot will be able to increasingly take work that you give it and finish it for you, kind of go that last mile in the background. This is an important evolution of Copilot, where in the past it’s really been a one-to-one relationship between the chat with Copilot and what Copilot can do, where it can start to be, you can chat with Copilot and then send it off to go complete a workflow in the background. And this is how we think we’ll see a huge, even a bigger increase of the productivity benefit and ability to kind of free people more of that drudgery. Then you start to kind of be able to focus and have longer periods of time where you focus on the hard part of the job, you know, planning for the future, doing budget, doing analysis, doing strategy—the parts that we all love to do, not the parts we don’t. 

MOLLY WOOD: Right. Say a little more, if you would, about the background operations and how you might take best advantage of that compared to the kind of real-time interaction that we have now.  

CHARLES LAMANNA: First is, Copilot today, since you’re talking to it, it can take, kind of, do actions and take steps in response to your requests, but it’s very one at a time. So, say if you want Copilot to help you along like a 10- or a 15-step process, you’re going to be sending 10 or 15 messages to Copilot. Get the data from the dashboard. Put the data inside of an email. Send the email, you know, so you’re kind of guiding it step by step by step. But if it’s something you’ve done multiple times in the past, and you have good examples, you can start to go to Copilot and say, Hey, every Friday at 4 o’clock, go to this dashboard, pull out the data, format it in the right way, and send the email to my boss. And you configured it, you’ve arranged and reviewed exactly what Copilot is going to do. And then you can kind of let it just run that task automatically each Friday. So you can really free yourself, and this really stays true to our principle of, like, a human is always in control and Copilot augments the person, because a person is still configuring and setting this up. But they just don’t have to be there for the 33rd time where it’s done these five steps asking it along the way. So now, that’s just one example. Well, you can imagine the typical office worker has 20, 30, 40 things like that they do every month, and this will make it so everybody has the tools and the capabilities at their fingertips to automate those parts of their own job. And that, to me, is what personal productivity looks like this decade.  

MOLLY WOOD: That’s such a game changer. Like, you could imagine how it changes people’s happiness and jobs and, of course, springboards them into their own creativity. On that note, let’s talk about the copilot-to-human split. You mentioned that there has to be a human in the loop. Now humans have the opportunity to do much more, much more fulfilling work. Talk about that split and how the tools and the humans work together.  

CHARLES LAMANNA: Well, we’ve always thought with Copilot, we should have computers do what computers are good at, and we should have people do what people are good at, and what people enjoy doing. People are great at creating ideas. People are great at long-term planning. People are great at collaborating and working with other people to complete a task. We don’t want to change any of those things. People are able to, you know, synthesize a hundred documents into a bigger document or read through a bunch of knowledge-based articles to find the right answer. They can do all of those things. Computers now, with the magic of generative AI and these new models, are able to do those things very well and can do them on behalf of the person. So we kind of view, like, if there’s a pie chart capturing the work that you do each and every day. In the past, a person had to do 100 percent both the monotonous, repetitive, mind-numbing tasks, as well as the creative, exciting, collaborative tasks. We’re having Copilot take up more of that pie chart for more of the mundane tasks and make it so people can spend more of their time each week on that creativity, that brainstorming, that collaboration with other people. And the best way for that to work is you, of course, need great technology, amazing AI models, there needs to be responsible AI filters and guardrails. You need all of those things, but user experience and change management is just as important. Because how can we take all that great tech and expose it to a billion people on Earth in a way that it makes perfect sense to them and they trust it to go take actions with them and for them. And then how can we make it so that you go educate and train and skill up the entire world about how to use these tools to be more productive. And if we think back to, there was a time when a typical office did not have a PC on the desk. You know, people wrote memos by hand and they had typewriters, and then PCs came and all of a sudden every single office worker had a PC, you know, a desktop and then laptop. The same type of thing is going to be true for Copilot. We’re going to go from a world where today most desks and most workers don’t have a Copilot to help them get their job done. But a few years from now, everyone will have a copilot to help them get their job done more efficiently and faster, and we’ll wonder, how did people ever work before they had an AI kind of copilot that could help them complete tasks more efficiently? Just like I now wonder, how in the heck could you run a large team without a computer, without email, without Teams? I can’t even fathom life without those things. So the same type of progression will happen through technology, through user experience, through change management.  

MOLLY WOOD: You have read my mind with the change management remark because you have, of course, been developing these apps and helping businesses adopt them, and I wonder how you think about where leaders should even start. With inventing these tools and deploying them, you know, in the right way as soon as possible.  

CHARLES LAMANNA: Yeah, so I think there’s three things I’ve seen work really well. The first is, find applications which use generative AI and produce results quickly and get those deployed. Because that, like, the good news is, every technology company has woken up and is building and shipping generative AI capabilities, so you don’t have to build everything from scratch. And this is where I always start, because so many companies and customers I work with, the first thing they do is they go and they have a team of devs start building stuff internally. That’s great. But that has a long lead time, you have to train folks, and they can, you only have so many devs on staff. But there are so many great apps out there. So many great copilots and AI functionality that you can just get deployed with a click of a button. Go look at apps first, in addition to the low-level infrastructure. The second thing is really understand the outcomes and business case for all of this generative AI technology as well. I’m a technologist. I think I could spend all weekend playing with all the different copilots and AI things out there, but that’s not what makes the gears turn for a typical enterprise or workplace. Instead, the investments in generative AI tools all center around this idea of, how are you going to improve customer experience, or improve the revenue per salesperson, or reduce the average time that a customer is on hold before they get in contact with someone in your contact center? What is the business case? So, every customer I work with it’s, think about it, what are the three, four metrics that matter most, that you want to move the needle on, and how could we apply AI there? And this keeps us grounded in the real value of technology and not just the hype cycle of technology. There’s always hype cycles, things going up and down, but if you produce business outcomes, it will never go away. I mean, that’s the beauty of these things. And then, the last part is, really focus on engaging your co-workers, your colleagues, the workforce, and make them part of the AI transformation. Because the most successful deployments we’ve seen are where the end users, and IT and tech resources, work hand in hand to get the technology rolled out. So those are probably the three, I’d say, lesser known but super critical components of successful generative AI adoption right now. And we’ll all learn a lot six months from now that might be a different list, but that’s kind of what we’re seeing right now across our customer base.  

MOLLY WOOD: This is a good reminder that Copilot actually just launched in February of 2023. So in a little over a year, what else have you and your team learned from the business using this technology?  

CHARLES LAMANNA: One of the things that we’ve really noticed is it’s a rare time where it’s a piece of technology that improves the actual quality of business process. And what I mean by that is your sellers sell better. They can spend more time with customers. They generate more revenue per seller. Or your customer service reps. They can talk to customers, deliver a faster resolution, spend less time on hold and more time helping customers—showing up in all the metrics that matter. Or for finance departments, you’re able to improve job satisfaction and save like 30 percent of the time it takes to do key financial processes like variance analysis or reconciliation each month and each quarter. So you’re seeing real business outcome in addition to the productivity benefits. So the throughput: more deals, more customer service cases, more financial activities that can run through the user. And this combination of more value, better quality of experience, and better productivity and lower operating costs are a rare combo in digital technology. I feel like you usually have to pick one. Here, you kind of can get both with AI, and that’s why at Microsoft we look at Copilot generative AI and go, Oh, wow, this is something different than past changes. This is a new big paradigm for how we think digital technology will be applied in the workplace. 

MOLLY WOOD: And then finally, I mean, I feel like you probably have 10 to 1 million answers to this question, but how are you using AI in your day-to-day?  

CHARLES LAMANNA: The first is, I think I probably get 300 emails a day and 200 Teams messages a day, so using Copilot and the Copilot chat, I can really quickly get caught up. Being able to go to Copilot and say, do I have anything that’s from a customer? Do I have anything that seems high priority? Do I have anything that requires an action from me today? And it gives me the answer right away. It’s game changing. And then I would say in my outside-of-work life, my favorite thing is, I love the image generation capabilities that are out there. I use those to generate pictures, really for any occasion, for the many group chats that I’m in with friends and family. And I think I always, kind of like it used to be, you’d send GIFs, at least I used to always send GIFs in these chats. Now I can create a tailored image and it, I don’t know, to me, it certainly drives endless amusement. Hopefully the other people in the group chats feel the same way. The thing I would say, which is kind of underrepresented a little bit with generative AI, is it really unlocks creativity. Because in the past, just like we talked about the programmers earlier—oh, I have to learn how to write code to participate in AI—I’d have to know how to be a visual designer, how to open up Photoshop and, you know, sketch out this picture, do the layers. I couldn’t do that. No matter how much time I spent, it was impossible. It was completely inaccessible to me. But with GenAI and the ability to create these images, I can be almost like a quasi mini designer and create an image which exactly captures what I have in my mind in a way that was just impossible in the past. And this is true for images, music, videos, but also automations, applications, dashboards, data analysis. We should just take the same frame of mind and apply it to all parts of our lives where things will just become accessible to everybody.  

MOLLY WOOD: Way to bring it back to work. Charles Lamanna is Corporate Vice President of Business Apps and Platforms at Microsoft. Thank you so much for the time today.  

CHARLES LAMANNA: Thank you for having me. 

MOLLY WOOD: Thank you again to Charles Lamanna, Corporate Vice President of Business Apps and Platforms at Microsoft. And that’s it for this episode of WorkLab, the podcast from Microsoft. Please subscribe and check back for the final episode of this season, where I’ll be speaking to Sal Khan, founder of the Khan Academy, about how AI is shaping the future of education and learning. If you’ve got a question or a comment, please drop us an email at worklab@microsoft.com. And check out Microsoft’s Work Trend Indexes and the WorkLab digital publication, where you’ll find all of our episodes, along with thoughtful stories that explore how business leaders are thriving in today’s new world of work. You can find all of it at microsoft.com/WorkLab. As for this podcast, please rate us, review us, and follow us wherever you listen. It helps us out a ton. The WorkLab podcast is a place for experts to share their insights and opinions. As students of the future of work, Microsoft values inputs from a diverse set of voices. That said, the opinions and findings of our guests are their own, and they may not necessarily reflect Microsoft’s own research or positions. WorkLab is produced by Microsoft with Godfrey Dadich Partners and Reasonable Volume. I’m your host, Molly Wood. Sharon Kallander and Matthew Duncan produced this podcast. Jessica Voelker is the WorkLab editor.