Transforming CX with predictable AI and workflow automation - Interview with Rebecca Miller of Pega

0:00So welcome to the Poxyx podcast. We're doing a bit of a special one today because this one's always linked to an announcement that's coming out via a friend of the podcast, which is Pegasystems. And with me today to talk about that is Rebecca Miller, who is a senior manager of product strategy and product marketing for Pega, particularly in the customer service and Pega sales automation space. space um so it's all quite exciting uh when this comes out they'll have just announced it probably the day before maybe 24 hours kind of max kind of later and so that's kind of pretty good so it's going to be fresh hot off the print printing press as it were um so wanted just to say rebecca welcome to the podcast how are you doing thank you so much adrian i am excited to be here i'm doing well enjoying this summer weather here in massachusetts it's high 90s today um but yeah really excited for this of this release we've been very um we've been very excited to share this with clients first at pega world and now really rolling it out on a larger basis kind of the future of customer experience and where the market is heading.

1:18Awesome. I mean, so can you just give us the big headline? What is the big headline? What has just been announced? What has just been released? Yeah, absolutely. So Pega has just unveiled a new agentic self-service offering. We call it the self-service agent. And essentially it is embedding enterprise workflows into any service channel. So this is really transforming how service is being delivered across digital voice and messaging platforms. And if we dig a little bit deeper, the new offering is really combining our PEGA blueprint offering, which is our online app generation tool that haven't checked it out.

2:01It's incredible. It basically gets updated on a two week basis where you can build out your own enterprise workflows and then actually use the power of our workflow automation and our predictable AI, so our agentic and generative AI capabilities to turn these workflows into agentic self-service experiences. Okay. Now, we're going to unpack some of that, I think. Okay. But before we get into the details of it, I mean, I wanted to sort of back up a little bit and just kind of get your view on sort of the other things that are going on in the market and the problems with them thereof and because your your your new announcement solution yes it's built on your kind of platform and it's an evolution of kind of what you you know what what you know what you've been doing and the the the the road that you've been following but what's the problem with the current solutions in the market from pegas perspective so we're seeing agents everywhere I mean, you look in the news, you kind of go on LinkedIn, every vendor is coming out with AI agents.

3:14It's a very, very noisy market. So there's obviously a whole lot of demand. I think that the last couple of years, the whole world was talking about generative AI, and now all of a sudden, it's shifted to this new tier of agentics. So it's really top of mind. But what we're seeing is that there's still a lot of shortcomings in terms of these AI agents that are out in the market. So if we just take a step back and, you know, we think about self-service over the last decade before the agentic wave chatbots, even though, you know, many organizations tried to use them were often really simplistic and limited, and they were really only able to answer pretty basic queries.

4:02So now with Agentec AI, we're seeing that chatbots are able to have more natural conversations. But the challenge is they're still being frequently deployed really in silos and in proprietary chat interfaces, really creating a disconnected customer experience. But at the same time, they're also really introducing a lot of unpredictability and risk for enterprises. So for regulated industries or really large brands that have so many different processes and they need to adhere to different compliance and governance, that lack of control that these AI agents in the market are basically introducing can be a huge risk.

4:55So how do you know that these virtual agents are actually going to follow your protocols, adhere to your policies and the governance? It can be really unpredictable and potentially detrimental to a brand if an AI agent actually goes rogue. Yeah, I mean, I guess if you think about it from a perspective of a large organization, and they're talking about deploying something which is particularly kind of customer facing, and then um legal and compliance get their hands on it and they go and you see this like oh it's like a 0.5 percent or a one percent chance that there's going to be some sort of like variety and the kind of response they're a bit like right hang on tell me more about that yes and we've been hearing that um you know from clients where their boards and um you know leadership teams really want to get on the agentic bandwagon and deploy these but then they have to pump the brakes because there's so much risk and um you don't want to have thousands of agents basically running around and making up their own um processes and so you know so is the kind of that your approach that kind of you talk about the announcement and it's like say it's fundamentally different but it's because it's leveraging kind of enterprise kind of like workflows i mean can you tell me about kind of that and let's unpack that a little bit for me because does that then take the variety out of it and the risk out of it yeah that's a great question and how i like to describe it is kind of what you said we're really taking the best of workflow automation, which Pega has been doing for over 40 years.

6:46And then we're combining it with the power of agentic AI to be able to transform every customer experience. So it is taking a lot of that unpredictability out of the equation. So it enables organizations basically to not only move beyond the limits of traditional chatbots and IVRs that were pretty scripted and not able to converse in a natural way, but also taking workflows and turning them into predictable and reliable self-service agents. But I guess if you're also going to tie it to this kind of workflows and you talked about this predictability, you can also, I guess, in many ways, um as if my if my understanding is correct you can also kind of like measure and and and observe kind of where things how things are performing across the kind of the way and you can break it kind of like down kind of more easy rather than it being a bit more black boxes as it were absolutely yes so that black box ai that we're seeing is exactly what we're trying to you know say is not the right approach for this.

7:56And really what, you know, we need to be focusing on is that fusion of workflows, which gives you that predictability. It allows you to ensure that the processes that are, you know, basically critical to your businesses are being followed. And then you're able to essentially instantly turn those workflows into conversational agents. And so you talk about this conversational agents, but you talk about this agentic application because it's all kind of like it can deal with more complex kind of queries. Can you help me understand how far the capability can go? How complex are we going to be able to get or can we get currently and where are we likely to be able to get to going forward?

8:45Yeah, and that's why I'm so excited about this is because these agents can really handle the same types of complex interactions typically managed by CSRs at really large organizations. So for example, they can answer detailed questions while dynamically outlining multi-step processes for customers, and then they can actually go in and resolve the inquiry without actually getting a human involved. So this could be managing something like a first notice of loss, which often did require a human CSR handling complex billing inquiries. And then these AI agents can really identify the types of requests that are coming in and guide the customer through the entire process and adapt to the conversation based on those customer responses.

9:43So that's really a unique piece of this, too, because I feel like, you know, oftentimes you could get super frustrated as a customer, you know, where you're sort of halfway through a process and the chatbot is kind of working. Right. You throw it a curveball and you break the chat bot. So, you know, that kind of goes away in this new paradigm of agentic self-service and the self-service agent. And what's really unique at Pega is, you know, I think you've probably talked to Don Sherman and some of our other, you know, members of the team about our center out business architecture. And this is really core to the self-service agent as well, where you're basically able to orchestrate those service workflows from an enterprise core and then use that semantic understanding that is so, you know, kind of baseline critical of agentic AI.

10:49So you can really ensure consistent, scalable service across all channels. So it's not like you're just deploying this in a single proprietary channel. Yeah, so I think that makes sense, particularly if you're a large enterprise and you're operating on different continents and across different borders. You might be making technology choices that span those different territories. But actually, what's interesting to me, if I think about the communication preferences of different territories, they're often not the same. like for example you could have like an international organization that is operating in both say europe and kind of the the us and so people have maybe using web-based kind of like chatbots and that's kind of one you find that's kind of one kind of thing you can do that sort of deployment but there is a complete difference between the popularity of say whatsapp as a channel in europe um and whatsapp in the us which is not very popular and actually it's more messaging and with the advent of things like rcs and all that sort of that that sort of stuff then that becomes almost like like not necessarily it has the same capabilities of kind of whatsapp so you could almost like as you're saying the center architecture seems to me is that you can decide at an enterprise level and a core technological level this is how we'd like to do things and then you choose then to deploy in whichever kind of channel you have the same capability regardless of the channel as it were as long as as long as it's got that that sort of the the the capability to be able to take those sort of interactions yeah i think we're really trying to make the channel sort of less of the focus right it's really just about that core capability and um the other amazing thing with these self-service agents is that they can really dynamically um shift to different languages as well.

12:50With Hega Blueprint, you can actually go in today and build out a workflow. And then you can preview that workflow and turn it into an actual voice or chat self-service agent right there. And you can actually call your Blueprint and have a conversation with this self-service agent that's following, you know, the workflow that you just created. But if you start speaking to it in a different language, if you start speaking to it in Turkish or Spanish, it immediately picks that up and it starts engaging with you there. So it's pretty incredible to just the multi-language support. You know, as we're talking about large brands that you know work across the globe where there's lots of different cultures and languages yeah no absolutely i i mean and i think they also the other thing i mean what i will say is a bit of an aside i saw some of the enhancements to blueprint that were announced at the uh the pegaworld this year you know from the the main stage and i'm particularly i would point people in the direction to look up the uh the demo that kerem did about the ingestion of uh different data types to try and move from a legacy uh technology stack into sort of like reimagining things in the blueprint where he did this like a video a talk through of a video of an old cobalt based credit management system and then took that and and chucked it into blueprint and ingested it and then it went this is what it looked like in a modern sort of like uh workflow kind of a modern kind of genetic sort of like stack and i was a bit like that's cool and just because yeah just because it's like you know that i think that it's interesting that the current level of commentary sort of suggests that like everybody's on the cloud and everybody's doing this and everybody's doing that you're like going no actually there's a lot of people that are still working on legacy applications.

14:58I mean, I wrote a piece about this where it talked about the idea of, what is it that only according to Goldman Sachs last year, that only about 30, 35 % of our enterprise workflows were in the cloud as of the summer of last year. And that's only expected to grow by 15, another 15 % over the coming three years. So up to about like sort of like 45, 50%. So there's still like a lot of work to do. And also another frightening, another two frightening snacks is like 70 % of all enterprises still rely on old legacy applications, including things like COBOL, written on COBOL, which is like COBOL is like 60 years old.

15:38There's like 250 billion lines of COBOL code still in use on a daily basis. That's like mind blowing. It really is. It really is. And what's so great about Blueprint is that it takes so much of the time out of having these discussions, and it allows you to actually just visualize what's possible, which has been really transformative for a lot of clients and prospects, so they can get an understanding of how you can actually transition to a more modern infrastructure without basically redoing everything yeah so um this is a new release pega self-service agent it's like a new release but as it always happens you'll have been having sort of like a close core of people kind of trying out and they can be in the wild as it were over the last kind of like few months i mean um what's been the feedback so far and what are people kind of telling you about kind of like what they like about uh this this new functionality no we've had great feedback so far and we continue to hear from our clients that they are looking to automate more and more out of the contact center so that trend is not going to stop i think that you know the reality is just based on our conversation on legacy systems that are still and play.

17:12It's going to take time for businesses to really adopt this new agentic approach. But, you know, we've been hearing on a fairly regular basis that there is really this need to switch to self-service and make that sort of the primary spot for customer service. And in some cases, we hear leaders that are actually planning to reduce some contact center head count. So at the same time, kind of like we talked about, we work with some of the largest brands on the planet. We also hear the concern that they have around governance and the control of AI. um so pega's approach of really combining predictable ai with workflow automation has resonated really well with our clients and many are starting to raise their hands to become early adopters of this and are already experimenting on pega blueprint as well as with our other gen ai and agentic capability.

18:21So we're really optimistic and excited about the prospect of self-service agent. And it's been great to hear that early feedback from clients. It's been awesome, you know, at PegaWorld to see people actually going in at the innovation hub and building out their own blueprints and seeing it come to life. Awesome. And when is it going to be in general release? So it is coming to Pega customer service and Pega Infinity 25 in the early fall. So look out for that. But it is available now via Pega Blueprint. So if you haven't checked it out, go to pega.com slash blueprint and you can see the amazing capabilities today.

19:13Awesome. And so I wanted to just like zoom out a little bit and ask about some kind of just general kind of questions around your perspective on the kind of where this space is at right now, unless there's anything else you wanted to add around the new announcement with the self-service agents. I guess the only other thing that I would say is just, you know, aside from the governance and control piece, there's also this benefit of really reducing the amount of development time, which we know is so important as well. So really being able to quickly build a workflow on Blueprint and then instantly transform it into a self-service agent and deploy it on your channels.

19:57It's basically taking months of work and, you know, moving that to, you know, days, which is really incredible. Yeah. I mean, I think I would sort of echo that. the other the sort of thing around um if anybody's ever done any sort of process improvement work where you end up having to map the kind of what is the as is and then the kind of the to be as it were and you find it's kind of hard to kind of get kind of like started because you end up with all this documentation all these different sort of things it's like going haha it just got easier because you take all that sort of stuff and you upload it and go tell me how it is yeah absolutely And then it goes, and then you can then start reimagining of how you would like it to be kind of like, and with all these new kind of like, you know, features and things.

20:45And it just fast tracks that whole kind of process because you can, when you see things, it's, it just, it creates momentum, it feels like. Yes, totally. I mean, it basically instead of, you know, having a white paper or a PowerPoint presentation where you're trying to understand a concept of how something will work, basically Blueprint allows you to actually go in and, you know, try it yourself and actually, you know, take that and put it into the Pega platform and start, start working from it. the other thing that i love about blueprint is just how it's able to connect it and business leaders so really being able to collaborate and foster that design thinking is really really fantastic and so that's brilliant thank you for that and so i just wanted to kind of say because i'm normally i try really hard to get to pega world and i have been there for the last few years but this year i couldn't go because i got floored by a bout of sciatica which was just really painful it's the first time for me and it was a bit well it was a bit rubbish oh you're feeling better now oh yeah no it's much better now but unfortunately that meant now just like the idea of a long haul flight to vegas at the time was not let's put it this way palatable um and so i watched from afar and i enjoyed a lot of the kind of key notes but missed about missed the innovation hub and talked to lots of people and and all and seeing kind of what was going on on the ground and you know and so what i wanted to ask um was you were there what was your main what were the main highlights for you particularly from like a customer service side of things you know what what stood out for you yeah it was a great show um was it hot It actually, it was hot, but it wasn't Vegas hot.

22:49I actually think we had a little bit of rain while we were there, which was a unique experience. you know in Las Vegas in June where where there was some rain so it wasn't too bad but it was a really exciting conference lots of exciting amazing announcements demos client stories specifically on the customer service side of the house you know we talked about the PEGA self-service agent where attendees had the opportunity to build their own blueprint so that That was fantastic. The other demo in the Innovation Hub that was really a great experience and kind of is along the same lines of putting people in the driver's seat and allowing them to have a hands-on experience was our customer service simulator booth.

23:46So this is basically for a customer service rep or agent to be able to, instead of, you know, diving right into an actual customer conversation, be able to engage with an AI agent and practice using customer service desktop. So we had, I think, maybe six or seven different desks where people could go in, put on a headset, and actually get into the shoes of being a customer service rep and engage with all of these different customer personas. Some that were very polite and others that were not. and people got pretty competitive about it but it was really you know pretty incredible to see how faces lit up when they saw all of the great capabilities that make it easier for customer service agents today was actually I think think the most attended booth at Pega World so that was a really exciting one um nice you know such a wealth of client stories that's one of the great things about pega world is that most of our breakout sessions are also client are client led we had a keynote from verizon where they talked about orchestrating the future of customer experience with ai and essentially they're using our ai and decisioning as well as our customer decision hub to create what they call a unique version of Verizon for every customer.

25:32I love that kind of notion. And they're also using Blueprint to dynamically generate new case types where they're able to create automations. And they've seen some really amazing results, improvements in terms of customer satisfaction, average handle times, I think it was about a 20 % improvement. but you know if you didn't miss it we missed you adrian um you're just saying it you're just saying that no we did we did um hopefully you'll be able to come next year but you can see all the replays at pega.com slash pega world if you weren't able to attend um you know it's it's a pretty great set of client stories and i think that's the most powerful way to really understand how the technology is is really impacting and transforming um across industries awesome and i mean and then sort of like you because the thing about kind of pega world as well because you get all sorts of different people that are sort of immersed in the sort of the whole customer experience customer service and the like world and you also get like a pretty good view of kind of where the industry's kind of heading as it were and it's not just about so the technology is just generally but you get a general feel for where people what's what people are seeing and hearing on the ground and what where the what they're driving towards and stuff and i want to do and get your perspective on kind of like with it with all of that what are your thoughts on where the service industry is headed and and are we just at the beginning with this agentic self-service tools or what's the future going to look like?

27:20Yeah, that's a big question. And I do think that we're just at the beginning of all of this. So, you know, based on the conversations that we've had with clients and prospects, as well as partners, it's pretty clear that there's going to be this significant transformation as well as consolidation of the front office. We're already seeing that a little bit, but I think that's going to accelerate quite a bit in the next three to five years where, you know, contact centers and sales teams are still going to exist. but that being said it's definitely going to become a much smaller function where some of these roles are going to be much more in the knowledge worker type of space where CSRs are handling the most complex or emotionally delicate types of inquiries and really the front office is going to be flipped on its head.

28:23So that's one thing that I think is really going to shift and agentic AI really plays, you know, a pivotal role in that. And I would say, you know, looking forward, even in, you know, five to 10 years, the vast majority of issues, whether they're complex or not are going to be self-served and we'll see more and more proactive and preemptive service um playing into that as well based on statistical ai and risk modeling yeah i mean i think the thing that i mean so i i i think i agree with with that but i also kind of think people can may listen to that and one of the quick conclusions that they'll they'll they'll jump who's like well there'll be no jobs left for people yeah i'm a bit like going yeah but here's the thing so these systems all need monitoring and managing and and you know at the time and new knowledge needs to be you know curated and kind of quality approved and all these different sort of things but also the other thing that we need to kind of bear in mind is that the number of products and services that most kind of brands kind of release is not going to shrink they just keep adding new things and when you add new things there's a whole raft of new complexity and new questions and all these different things that kind of that you have to add on to um you know this kind of like space so that they ever there's an ever evolving complexity um that that exists and And so I actually think that the role of people is going to, one is going to shift, that it's going to be there to deal with the variety and the complexity that is going to exist in this evolving sort of like space.

30:15And so I'm not, I do think that the technology comes with choice and we have to be very cognizant of the choices that we're making around the technology, you what it enables us to do but i think there's there's a lot that can you know there's there's a lot to be done by human beings if they kind of if they grow and evolve their kind of like role and get the right support in doing that i completely agree um i think that just the types of roles are likely going to shift and uh i think you know you need to be curious as someone who's an employee or customer and be open to those shifts. We're really encouraged to use AI tools to augment our daily tasks, our work, rather than replace them.

31:18So I'm definitely in the camp that not everything should ever be completely automated. I think there are certain things that do require a human touch that being said um you know i think that there is going to be a shift in terms of the type of work that a csr is going to actually be doing when they log in for the day yeah no no indeed indeed um so rebecca that's i mean that's it for my um some of big questions If I may, anything else? Well, anything before I go on? Anything else you want to highlight before I kind of go into some quick fire questions? I think we covered it. Let's go into the quick fire questions.

32:05Quick fire. I've got three of them. Okay. First one is we've had a bit of a rummage around. We've talked about this, that, and the other. I want you to boil it all down. And how I'd like you to do that is to complete the sentence, Rebecca. And the sentence is, if you want to improve your customer experience, Rebecca says, do this dot, dot, dot, complete that sentence. Start with picking one or two workflows that have a potential to be automated. Maybe it's part of a claims process and begin infusing AI and automation. Don't wait. Full stop. Thank you very much. um second one is yeah i mean this is a punk related one so i couldn't not because the punk cx podcast so here's a question which company or brand do you think is taking a more punk approach to customer experience and why this is my favorite question um and i'll do it from kind of like a personal experience so i i really like the brand ikea you know aside from having to build my own furniture where i often will need to recruit my husband or friends to do that i'm just not the best at it i think that they're great it's fun to go to their stores um do you like the meatballs it depends on the day what about you never sure i mean i'm never sure about going into a furniture store and then sitting down in a cafe it just feels like incongruous yes yes um but you know sometimes they hit the spot but what ikea is doing if you know we stay kind of in like this ai space is that they're blending augmented reality and virtual reality to help let customers now visualize furniture homes before buying it.

34:09So basically, you can get a sense of what your furniture is actually going to look like before you actually go into a store and buy it. And I think that this is really great just from like a convenience perspective. It also is saving some waste if you have to return something where you get a sense of it. So that's super exciting to me. And also their app offers sustainability scores. So you can get a sense of the type of furniture that you're buying, how sustainable it is, which, you know, personally, that's important to me. So I love that we're starting to see these consumer brands embrace augmented reality and VR to really be able to transform the customer experience and the buying experience in a pretty immersive way.

35:11Nice. Nice. Well, I thank you for that. I can't remember if IKEA has been on the list before, but it's now firmly on the list. So thank you for that. yeah um and final question before we wrap up i've been doing this for a little while now because the world for a little while now or probably for a little longer than that has been slightly weird yeah and kind of odd and news is dominated by kind of just like bad news it feels like so i've been asking people to you know to share with me a good news story before we wrap up so rebecca can you tell me tell me a good news story tell me something that you've seen in the in the last week which is like interesting positive or exciting it's made you smile you're going ah the world it's not all doomed it's made a it's kind of it's made me smile yes so i saw a story a few days ago around how the la county so los angeles is using now ai to proactively identify and assist individuals that are at high risk of homelessness.

36:18Okay. So essentially, they're able to use this technology to prevent them from actually losing their housing. And it enables them to be able to support with things like rent assistance and mental health care before it gets to the point where you end up, you know, losing your house. And I saw a stat that it reduced shelter admissions by 71%. Nice. That's a really nice story. I have a lot of family in LA. So I keep up with the news over there. And I know that they've had challenges in the past. So just hearing those stories of how AI is really, you know, kind of contributing to the greater good really makes me optimistic and excited for what's to come.

37:08Awesome. Well, thank you for sharing that. That's a really nice story about how folks are applying technology to look after people. Yeah. Which is nice. So thank you for that. I just want to say thank you for sharing your time and your insight and your expertise with me today. Congratulations on the whole announcement around the self-service agent. That sounds really exciting. People should definitely go check it out and check out it, you know, wait for either join the queue to get kind of like to talk to somebody at PEGA about it or check it out on the the blueprint page which is PEGA.com forward slash blueprint I think and you can see it all kind of like there but finally yeah Rebecca thank you very much it's been great thank you so much Adrian I really enjoyed our conversation and I love ending it on a positive note you're welcome thank you well that's another interview in the bag I just want to say big thanks to Rebecca Miller for spending time with me telling us more about Pega's workflow-powered agentic approach to customer self-service.

38:10I hope you enjoyed it. I know I did. Now, but before I go, there's a couple of things I'd like to mention. First of all, please consider heading over to iTunes, Spotify, or whichever podcast platform you choose to use to listen to these interviews on and do leave a review. It all helps. Thank you. Two, if you have any comments about any of the interviews I do or have any suggestions for topics that you'd like me to explore or people that I should interview, then feel free to send me an email to hello at punkcx.co.uk. Finally, do head over to adrienswinscode.com if you want to find out more about how I work with clients in this space.

38:48Anyway, that's all for now. Thank you for listening and do tune in again soon. All the very best. Bye-bye.

38:58Let's go.