Latest / The Wilmac Wire / Behind the Build | Two Products, One Vision, & How Agentic AI is Changing Everything

Transcript

Behind the Build | Two Products, One Vision, & How Agentic AI is Changing Everything · The Wilmac Wire

• speaker-0: Good morning, good afternoon, good evening, welcome or welcome back to the Will Mac Wire. Thank you so much for tuning in. I hope that while you're listening to this episode, you're enjoying some nice summer weather, bright blue skies, not too many clouds, some sun, feeling that vitamin D on your face. ⁓ the weather has finally turned here in Rochester, New York. And as you can tell, I'm I'm a little bit excited. I'm sitting In our one of our conference rooms in the Wilmack Technologies headquarters based in Rochester, New York. We're up on the tenth floor, so we have a pretty sweet view of the city. And that's what I'm seeing: bright blue skies and sun. We just wrapped up our company-wide half-year meeting a few weeks back where we, for the first time since I've been here, so at least in five years, we've had we had everyone in the company in our Rochester headquarters to enjoy some. Some vision alignment, some team building, and then setting our goals as to what we want to achieve in the second half of this year. So for today's episode though, let's stay on track here for what w what we're here for today. You know that every so often we like to turn the mic inward and check in on what's happening inside of Will Mac. We call these episodes behind the build and they happen quarterly, so we were due for another one. Every behind the build episode, we sit down with Bob Pease, our VP of engineering, to talk about where our products are, what we're seeing in the market. And more and more frequently with these episodes, they've become a conversation about AI, where it's showing up in our products, where it's showing up in our industry, how it's showing up in our business and how we're utilizing it day to day and trying to pioneer it within our region. So it's a fun one. I always enjoy these episodes. Here's what you can expect. To get out of today's Will MacWire episode. You'll get an update on both Will Mac Continuity WFM and Will Mac Continuity Replay, including what's evolved, what we're building, what's next, who's contributing. We also, of course, are taking a look at AgenTech AI and how it's showing up across both products with our friends Mac and Archie. Those are the names of our AI agents for WFM and replay, respectively. And like I mentioned before, a broader conversation on AI, what we're seeing, how it's being utilized within our organization, and where we think it's going. So really exciting conversation today. I will also do a little disclaimer that we're kicking things off a little bit differently. You guys know I love my icebreakers. So instead of this or that, I am going to force Bob and Steve to play a round of on point or off point, I'm calling it. So to get a better idea of what that means, I'm sorry you're gonna have to have to keep listening. So thanks for tuning in to today's episode of the Will MacWire. Let's get into it. Welcome to the Will MacWire. I'm Emily Miller, joined by my co-host, Steve McDonnell. On this show, we talk with the drivers, the leaders, innovators, and decision makers pushing customer interaction technology and AI forward. And because we're proud to call Western New York home, we also spotlight the drivers in our own community, the people and organizations shaping growth and innovation right here. Let's get into it. Welcome listeners back to the Wilmac Wire. Thank you for joining us today. It is a repeat guest today. This is our quarterly behind the build update. So we have just headed into quarter three at Wilmac. So I'm here today with my co-host, Steve McDonald, and then Bob Pease, Wilmax VP of Engineering. Bob, welcome back. speaker-1: Happy to be here. Excited as always. speaker-0: ⁓ yeah, we we love we love having you on the podcast, especially when we're at our nice little round table at Rock Box Studios. ⁓ so before we get into kind of our topics of discussion today, you guys knew an icebreaker was coming and I couldn't do what we'd done in the past and just be r repetitive. ⁓ so I have a new one for you today. I have five prompts all revolving around AI, and I'm gonna ask you and Steve to tell me if you think it's on point. Or off point. So on point being you agree, it's relevant. And off point being, ⁓ you disagree, it's irrelevant. All right. So we'll we'll work through this in real time. Okay, starting with the first one. AI will replace more jobs than it creates in the next five years. That's a loaded one. I didn't realize. speaker-1: I think yes and no. ⁓ I mean you we can already see where AI is replacing jobs. People are companies are doing mass layoffs for AI, that sort of thing. But we can also see, I mean, Microsoft just announced they're rolling back their clawed usage because they're spending a lot of money, right? So I think we're still kind of finding our footing here in the industry, but I think the thing I feel most strong about is we're gonna see a lot more AI focused jobs coming out, right? Roles are gonna shift and it's gonna be more focused on AI specific roles. speaker-0: I didn't realize I had that one. speaker-2: Yeah, if you had asked me this even three months ago, maybe four months ago, I would have said definitely AI will replace more jobs. Now with kind of everything shifting, I'd there's definitely going to be ⁓ layoffs, but I think it's gonna create just as many, if not more, jobs. I think there's gonna be new openings, new types of jobs. Yeah. That AI needs. So I don't think it's going to be a complete displacement. I think it'll be a displacement in certain jobs. Yeah. And I do think people coming out of college, as we've discussed, you need to know AI. If you don't if you don't know how to utilize AI, you are behind. And no matter what role you're going into, you have to be able to learn learn how to harness and utilize AI. But I think it should create more than displace. speaker-0: I agree. On point. No, I'm kidding. AI will replace more jobs than it creates in the next five years. That's off point is our is our sentiment. All right. ⁓ next one. Companies that haven't started their AI rollout by twenty twenty six are already behind. I want to start with speaker-2: I think that's on point. I mean, I don't care what yeah what industry you're in, if you're not utilizing it or beginning to utilize it, I think you're behind the eight ball. I I obviously some companies are way further ahead. Yeah. Everyone should be dabbling in it. I see a lot of large companies that aren't technology companies really being forced to use Microsoft Copilot, and I think that's gonna hinder a lot of these companies. They should be and and they're doing it for compliance, procurement, partnership reasons, I don't know, but not using one of the the clawed open AI or even Groc or what's the one sort of sales guy perplexity. ⁓ Not using those and just sticking to Cold Pilot, you're gonna get behind the eight ball. ⁓ so understand that not everyone might be as far as long as some tech companies, but you should be having some sort of plan at this point. speaker-1: Yeah, I think that's entirely on point. ⁓ it just it it happens so fast and if if you're not adopting it, it's gonna be kind of forced upon you for lack of a better term. Like every software that you're using is working on their own AI initiatives, right? So like if you're not getting up and running internally, you're you're definitely behind. speaker-0: That's what I was gonna call it, like all the subscription software that we use, like there's just AI capabilities being added to it. So it's being pushed in our faces and we're in it every day. so yeah, I I agree that that is on point. Okay, next one. Most companies saying that they have agentic AI don't actually have agentic AI. I feel like this goes a little bit to the conversation we were just having with the different types of speaker-2: We're gonna dive into some of the agents we've been building. Yes. ⁓ Mac and Archie later. I think it's tough. I mean, some of this is more like verbiage and theory. I think there's a lot of companies out there that are have some great SaaS solutions or software solutions that are probably just building a chatbot or a copilot and saying they have agentic AI. I think that's I think that happens a lot, is what I'll say. But then y you know, a true there's true SaaS platforms that also are building true agentic AI agents. And then there's just pure AI companies that never even built platforms that are just purely agentic AI. So I think it's both. speaker-1: Yeah, I don't know that I have anything I can even add to that to be honest with you. speaker-0: It's a marketing thing. That's honestly what it is. And like yes, there are the clear definitions of it and we'll talk about those different kinds, but ⁓ it's all about your perception. So ⁓ next one. I think Steve will like this one. Vertical AI is the future. On point or off point. Yeah, you're gonna say that. speaker-2: Our future. I think for when you look at the tech stack of AI in general, where we play is the application layer. And I think the best place for application layer is niche verticals. I think that's the toughest for the large language models or future LLMs to adapt to and people just to be able to build it quickly, either themselves or a small company coming and starting up. I think vertical AI is specialized. They have specialized workflows, specialized compliance, security to get into these places. If you're verticalized, you have distribution models that other people don't have. so I think in the application layer of AI, which I think will eventually become one of the most abundant industries in the history of the world, I think vertical AI is a good spot to play in if you're speaker-0: Yeah. And I mean for our listeners too, Vertical A, I like you said, it's specialized, it's indu industry specific. speaker-2: It's very specific industry. So Willmac builds continuity WFM for PSAPs or 911 centers. If you want to even broaden a little, you'd call it public safety. But that's a very specific industry. We build continuity replay for regulated industries. That's anyone that has severe compliance with a lot of data, which that's a little broader of ⁓ vertical, but it's still a vertical. It's not horizontal. A horizontal would be like your open models right now ⁓ or like a payroll company. You can basically sell to any company out there. Right. Vertical companies are selling and building to specific industries and they build specialized products for that those industries. speaker-0: Yeah. Bob anything to add that? speaker-1: Not a ton. ⁓ I'll I guess th what came to mind for me is literally anyone can build an AI product using an LLM. It's just it's not that hard. It's so readily available to us. That does not make a great product. The people that are really making a splash out there in the industry are the folks that are focused on specialized products on niche industries. speaker-0: Yeah, solving a very specific problem. Yeah. Okay, last one. Bear with me. I don't think I should have done five, now reflecting back on it. But last one. Okay, the best AI products being built right now are coming from small companies, not big tech. I think Steve has chuckled at every single one. speaker-1: Well speaker-2: I think is it products or applications? I mean the LLMs are now not small companies at all. Companies in the world. I wouldn't call I don't know if you call that they are products, especially for the consumer and stuff. Are they enterprise products? I I guess they're taking over software engineers. ⁓ so I don't know. That's a mix. I think smaller companies have way more access and ability to build awesome speaker-0: No. speaker-2: applications now. But I don't think a smaller company could go out and get into the other kind of s ⁓ tiers of AI. I think the only part of the stack that smaller companies can play in at this point due to capital really is applications. But as I said earlier, I think it's gonna be massively abundant. So a smaller company could build something awesome way easier where you might need Even five, six years ago, you know, you need thirty software developers to build an awesome product. You need two or three AI engineers now that can build this awesome product. So it it opens up the world to smaller companies. But I still think some of the larger companies have a good grasp on what's going on and a in a big advantage. speaker-1: I don't think that AI changes this from my perspective. I mean, even historically, smaller companies are usually more agile and creating cooler, right, more disruptive type technologies. And I think a lot of the big players today are helping the smaller guys really disrupt, right? When you look at anthropic or open AI, like they're adding the tools that enable those smaller companies to go out there and and crush it. Right. speaker-0: All right, awesome. Thank you guys. speaker-2: At same time they're trying to speaker-1: Take away that business? speaker-2: Let end users just build it themselves too. So it's it's kind of a interesting market going on out there and it'll be interesting to see how it plays out over the next few years. speaker-0: That's for sure. Yeah. Thank guys for humoring me. I know we didn't stick to the on-point, off point answer, but it's it's facilitated some good discussion. ⁓ all right, we're gonna jump now into kind of the the crux of our typical behind the build conversations and we're gonna hone in on Consume EWFM, little Mac Consumer WFM to start. So, Bob, I'm gonna let you get things started by sharing kind of what's changed and evolved since last time we spoke in February. speaker-1: ⁓ so much. I love WFM. Yeah. Working on it with the team because we're getting through so many problems. ⁓ even just on a weekly basis, whether that's a simple enhancement that we hear about, a major feature we have to release, or you know, the favorite my favorite part is the AI initiatives that we're we're running with. ⁓ I think the biggest and and probably the thing I'm most proud of with the team is ⁓ the abilities we've added to Mac. ⁓ So you can now add skills. And why that's powerful is ⁓ speaker-2: Wanna just bring it back quick for our audience. We're throwing out WFM Mac. Yeah. You wanna just give a quick thirty second overview of what WFM is and what Mac is speaker-1: So Continuity WFM is a workforce management tool built specifically for PSAPs. So folks run a 911 centers and how they have to manage scheduling and ⁓ people management of ⁓ of those centers. Mac is the AI agent that we built alongside Continuity WFM that really serves every layer of your PSAP, whether it's your individual telecommunicators, ⁓ your supervisors, your directors, et cetera. Back to adding skills. We have all these cool things we can do with Mac, ⁓ automating workflow approvals and generating requests and analyzing data. Skills allow a supervisor to build their own flow around all of those features to complete bigger, more complex tasks. So my favorite one to lean on. A lot of times when a supervisor walks in, the very first thing they do at the start of their shift is come up with a report. I need to know what's going on today. Is there anything critical that I have to be aware of? Any major changes that are coming? You can now build a skill with Mac that will go pull all the relevant data out of continuity WFM and produce that report for you to help you know what fires you have to put out right away, what issues am I gonna address during roll call before people even go sit down. And that has been really powerful and really resonating with our users. speaker-0: And skills y they essentially help you further customize too, right? Depending on the P sap. speaker-1: Correct. We did not invent the idea of skills. It's common in pretty much any AI platform out there. But yes, we yes, absolutely. We lean on a Claude example all the time. ⁓ a skill is just you write your own instructions that get fed into the agent so that when he runs his calls, he actually knows what to do. speaker-0: Okay, and what would an an example of a particular skill be? Like I know we've talked about it before, but okay, someone deals with a very heavy call. Is there a rule around okay, they have to have a certain amount of break time after that? speaker-1: Yeah, that's a great yeah, that's a great call out. So yeah, if if a nine one one call taker is on a really long traumatic call, oftentimes they'll get pulled off of that role for a small period of time, ⁓ which might mean that they don't get scheduled for call taking for a week. Right. ⁓ now if we think of the workflow of a supervisor, they have to go and mark all of that and and track it and make sure they're not getting scheduled there. Well, you could build a skill that would define those rules effectively within Mac. And then when that scenario comes up, hey, Emily had a traumatic call and I need to pull her off call taking for a while, it would just run that flow for you. Yeah. ⁓ and update the system in the back end so that you don't get scheduled as a call taker for whatever that defined threshold is. Right. So it's really flexible because in plain text, you can just describe what your rules are and Mac will enforce them for you. speaker-0: Yeah. Very powerful. okay, looking at roadmap, so ⁓ the kind of the product strategy has been very top of mind for us. ⁓ at least for sure over the past quarter it's always been there. ⁓ but house the team's thinking about kind of what's coming next. We're having a lot more conversations with PCAPs around it that you guys are being involved in. So how's that going? speaker-1: Going great. ⁓ first off, I'm pulling more of my team directly to those conversations, ⁓ which has been I don't know, night and day difference because now I don't have to go and have these conversations and then try to regurgitate the information back to the getting it firsthand. So ⁓ we're just executing way faster. ⁓ our roadmap is focused on really two areas. One, are there any big bang features that people are calling out ⁓ that we need to add to the product overall ⁓ for quality of life for a PSAP. So a great example of this one is a lineup feature. ⁓ one of the customers we're working with, they don't just assign all of their workers to specific stations. They actually get to choose on a daily basis. And this is a huge perk for them. ⁓ so we went and rolled out a feature and then come to find out other PC apps were also interested in this feature. So that sort of thing is where my mind goes when I talk about big bang features. ⁓ the other area is making Mac more autonomous. That's what's on our roadmap. ⁓ Mac started as really just a chat bot to help guide people through and understand what continuity WFM is. And then it became kind of a co-pilot and could answer a few more questions that were specific to your data. Now we're in agentic mode where it can actually go make changes to your systems based on your requests. And the next logical step there is fully auto ⁓ automating it. Sorry. Yeah. So that you can actually schedule Mac to go do things on a regular basis. Right. Every day, Mac, I want you to go and analyze all the time off requests and approve what can be approved and flag anything that needs further review by the supervisor. speaker-2: Can you discuss team development? ⁓ I don't want to steal any thunder, but we recently had our first like pure AI engineer hire. He's been with us not even a month and is off to the races. ⁓ pretty impressive. So can you talk about I I'm assuming most of the new additions we'll be making in the next, you know, three months to 18 months will be probably pure AI AI people. speaker-1: Hundred percent. Yes. ⁓ so night and day difference for a skill set on the team. So we hired a data scientist who has experience building out AI for large enterprises. ⁓ and just the skill set that he has and the way he thinks about these problems ⁓ is is why he's moving so quickly. And of course he's leveraging AI to develop his code just like the rest of the team. So ⁓ he's been doing awesome. ⁓ when he came aboard, we did our ⁓ breakout sessions during our mid year and the first one we just kind of went through and taught him here's the architecture of what we have today. And he had some ideas of what he thought he wanted to do. And we said, hey, will you take us through and teach us these architectures? And then from there, just all of a sudden we started getting to work. Like we all saw where the gaps were and where the alignment was and he was able to just go and start building. ⁓ and then they're coming out in all of our team meetings, like we're teaching each other things. ⁓ and that's been awesome. And I can tell you for sure We made a few mistakes along the way in developing our AI, but he's coming in and saying, ⁓ if you take this approach instead, that's what makes it better. Here's how you get your efficiencies and optimizations in your AI. speaker-2: Mm-hmm. Yeah, and one of the cool things, obviously he's working on Archie and Mac. Archie has a large customer base, so he's really running towards that. But his feedback kind of comes from a, you know, a Fortune two hundred, Fortune three hundred company. And just on how, and it's a it's good feedback for you and how you've developed your team, how he can take something, run with it, and we can be b deploying it at a customer site. within days, if not next week. Or the sales team could be coming in and talking to the product team and being like, Yeah, let's go and do this. And they're getting a real live example that's gonna go out, be shipped quickly. And coming from a huge company that he comes from, that's just not how it works over there. Everything moves slow, bureaucracy hits. You could be working on something and it could take a year or two to hit if it ever does hit. ⁓ so I think the excitement about working at a a company like ours, a small, flexible kind of push in the edges right now. he was very excited about that. And I think that definitely is something we need to take advantage of. And it's an exciting moment for Wilmack ⁓ to be there right now. speaker-1: Yeah, it really is. ⁓ and the autonomy ⁓ that we gave him to just go out and start experimenting building. Like I told just go build whatever. Like he and he would ask me, Can I use this technology or that technology? And I was like, Yes, go do it and then come back and teach us what you learn. That's the key thing. And that's been working really well. speaker-0: Yeah, I gotta give him a shout out a shout out if he's listening. Sammy Gole. Welcome to the welcome to the Will Mac pack. Sammy, we're happy you're here. And you're definitely a lighter on the office. Always always makes me feel welcome whenever I walk in and says hello. Sammy is the man. ⁓ all right, Steve, I wanna ask you a little bit about conferences and the speaking experiences that you've had at those. So we we're talking about products here and now we're talking about the building of the back end that thought leadership. But we're also trying to translate that to educational efforts ⁓ at these conferences and we're ramping up our conference attendance quite a bit. ⁓ so can you talk a little bit about your experiences in a customer? speaker-2: I think we've kinda taught told the audience, but WFM ⁓ workforce management has been around contact centers for twenty to thirty years. we have built it with AI and bringing it to the only call taking industry that we believe isn't utilizing it, which is crazy as they're the ones saving lives. ⁓ so there's a lot of thought leadership that goes around it because the people we're trying to pitch this to and the customer base has never even Heard of it, doesn't know how it works, ⁓ and doesn't have a budget for it. so we've been going to conferences quite a bit the past three, four months. Two months. Three months probably. And we have a heavy schedule coming up. ⁓ we're speaking. ⁓ it's exciting to get out there ⁓ and doing these speaking engagements. We've been kind of going back and forth. Are we speaking to higher level buyers when we do this and where they want, you know, some very high level value or are we getting in speaking to the actual users and they want the nitty gritty. ⁓ so trying to figure that out has been difficult but exciting and I think we've we've gotten there. We have numerous people on our staff who can speak, which is great. ⁓ and it's been going well. I think it's a it's a unique and small community, the PSAP world. So you go to all these different conferences and you're running into the same vendors and similar people. ⁓ so I think we're doing a great job on it. We're learning a bunch. We have a pretty large team, so it's not always the same three or four people going. I think we probably have twelve to fifteen people that are able to go that we split split time with, but we are traveling literally California, Texas, Arizona, Florida. So people are going next week. so all over the country. As again this platform, Continuity WFM, is new and we can get it across the country. speaker-0: All right, well. speaker-2: so overall pretty awesome. Mm-hmm. Yeah. Do you have some exciting news on the front as speaker-0: Yeah. Yeah. Well yeah, I just wanted to add to it too, like not only are we bringing a new product to market, we're bringing a new concept to market. So that's that's super exciting to talk about. The wellness piece has really started to resonate. And Bob, you just talked about it with the lineup feature, like giving ⁓ user ⁓ telecommunicators the opportunity to kind of pick their own pick their own seat for the day. That adds an additional level that if they're told it versus get to choose it, like it it definitely affects that wellness piece. ⁓ but yeah, on the conference side of things we've been ramping up so much and I I'm personally a little tired of figuring out furniture for conference for our booth. A lot of logistical effort goes into conferences. Yeah, I could probably start a whole podcast on what what tables to choose for your booth. but Haley Gunther is joining the team, a a local Saint John Fisher grad. ⁓ she actually starts the week this this podcast will launch. So ⁓ we're pumped to have her join the team and just jump right into the swing of things. Like Steve said, we've got a big lineup of conferences over the course of the next few months. We're in Florida next week. We'll be at the two national ones. So Nina twenty twenty six at the end of ⁓ this month in June, and then APO twenty twenty six down in San Antonio in August. So speaker-2: The marketing team's been holding the fort down, organizing, getting people together to go to these things and no one's really dedicated specifically. They have all their other marketing stuff to do. So getting Haley in here as a dedicated ⁓ events coordinator for us should really It'll be huge. Be huge for for Wilmack. speaker-0: Yeah, I'm pumped for it. These in person, I think, interactions, like Bob, I know you've been to a few conferences too. Like they're just so valuable. And even if ⁓ our our ultimate goal is awareness and we wanna generate interest in t WFM, but just chatting with people and building that community is really powerful. Do you have anything to add on the community? speaker-1: In particular because, you know, the 911 community, they are very collaborative. They're not competitive with one another. And just when you're in that environment, it's a totally different vibe. And and I don't know, the ideas that we come up with live in those start like very small out of a specific need with an individual and just kind of snowballs again and again. So it's it's so valuable for us to be there. It is speaker-2: Yeah, be and before we move on, I I want to ⁓ congratulate Bob, John, Ben, Chad, Evan, the kind of leaders of WFM. We have several pilots out there currently ⁓ with some of the largest nine one centers in the country. Are not they are the largest 911 centers in the country. ⁓ we officially launched in December ⁓ and we are doing pilot to to buy ⁓ program. But we received our first ⁓ full PO paying customer last week. one of the t three largest states in the United States. So exciting, exciting stuff. Hopefully we'll congrats a lot more to come. speaker-0: And you know. Congrats Bob to that customer. I can't wait to have you on the podcast. All right. Anything else to share on WFM and Mac before we move into replay? speaker-1: Yes. speaker-0: Yep, I love it. All right, sweet. So let's move into Will Mac continuity replay then, just to give less our listeners a little bit of context. We we've talked about this before, but the theme lately has for sure been in the public safety space. As Steve called out, Will Mac Continuity Replay vendor neutral compliance archive for highly regulated industry industries. So they're dealing with all their contact center data, their trading floor data, those communications that they need to archive ⁓ for a prolonged period of time. ⁓ so just as a quick brief as to to what Wilmac Continuity Replay is. So it's also been a part of Wilmax DNA, I guess I'd say for ⁓ the past ten years. ⁓ so Bob, do you wanna just start things off where by talking about kind of the updates? What's where's the product stand today? What's been the focus for the dev team recently and dig into that? speaker-1: Yeah, absolutely. ⁓ I'd say, you know, the the big there's there's not a ton of huge feature requests that have come up recently, ⁓ which is a great thing, let me be clear. ⁓ so we've been focused on strengthening our security posture, adding some new compliance programs ⁓ to our belt, which is awesome. ⁓ but for the dev team specifically, ⁓ we're rewrote the UI. So the original UI we wrote for continuity replay cloud. It's fine. Modern, looks nice, but ⁓ we started to realize it could be better. It could be even more modern, more compact. ⁓ so we have a whole new user experience that solves a whole bunch of I would say user confusion problems that we've had in the past. So we're launching that. ⁓ we're in progress of launching that now. Yeah. That's been going awesome. ⁓ and then secondarily building out Archie. So we Archie's in the product now. he's he can help with some of the simpler tasks. ⁓ we'll get into where Archie's going in just a minute, of course. But you know, he can help build out your dashboards and do reporting for you and quickly ⁓ find calls within continuity replay. So you can describe what you're looking for, it'll search your metadata for you and help automate that. speaker-0: I love it. Before we move on, I have to ask Steve to explain how Archie got named. speaker-2: And you want the honest truth? No. Archiving Archie. Yep. I was thinking about it and I came up with it. I was fooling around with Claude and OpenAI, throwing in certain things, and they gave me a bunch of options, and I liked Archie the best, so that's what I brought to you guys, and everyone seemed to like it. So we ran. speaker-0: I was to say, when we don't put a vote to the company as to what to name the AI, that went pretty quickly. speaker-1: I know it's kind of a silly thing, but I think it's so important that it ⁓ my gosh. Like AI in your products needs to have a bit of a personality. speaker-0: Sure does. Well, yeah, there's that whole we've had so many branding conversations about how we market it, maybe making sure it stands out in the product, has a little bit of different color scheme. I love that sort of psychology and thinking about that. ⁓ but Bob, I also wanted to ask, so you've been at at the company now what? Almost three years. So you came in, launched Continuity Replay Cloud. Now we're we've got a new UI, we've got Archie, like Development wise, how much different is it from when you started two and a half, three years ago to to today? speaker-1: such a big question. It's it's so different, but I I think the reason it's a big question, it's this has nothing to do with three years ago. This is like over the last six months and how much AI has changed and how much myself and the dev team are leveraging AI to do our jobs and shift away from some of the old approaches to software development, which was everyone get in a room and we do all this planning and everything gets scheduled and then you go fail a bunch of time. Everything just kind of moves really, really slow. I would say we've always moved fast by everyone's standards, but compared to where we are now, I'm like, Holy smokes, this is insane how quickly we move. So that's that's the biggest change I would say. speaker-0: Yeah, that's that's a great point. Like we've talked about being flexible and moving quickly. That's kinda where I figured you would you would go with that. But I mean, Steve, from the standpoint of when this product was built ten years ago. speaker-2: It's a lot different, that's for sure. speaker-0: ⁓ I did want to talk too a little bit, Bob. You mentioned about kind of like the future of of Archie and the agentic AI capabilities within it, because we're dealing with a lot of data as a piece of that too, which I feel like creates a really fun foundation. So can you speak kind of more to where Archie is going? speaker-1: Yeah, absolutely. So the next iteration of Archie is going to include a lot of in the weeds functionality that's really built around searching and analyzing your actual call content or any interactions, whether it's chat, email, et cetera. ⁓ so we we have transcription in the in the product and we're gonna be I don't know how technical I should get. We're gonna storing those in a vector database. And what does that get you? You can do things like semantic search. So imagine the really simple use case. I want to know anyone has sat talked about a jacket on a call. Well, what if they said coat? That matches, right? Because we it's the same thing. It has the same intent. So that's what to technologies that really Sammy's driving and bringing into the product are going to help enable us with. ⁓ and then to take that a step further, if we think about audits that happen, a lot of our users today, they just get an email from someone, whether it's internal or external. I need calls that match this criteria from this period of time for audit purposes. And it's like, All well, now they have to go in and search. Well, we can automate that whole process and like really cut that person out of it. So that they don't have to spend hours searching through. They can just say, hey, I need to look for those calls that say jacket. And then Archie will go and find it all for you, compare it against your compliance policies, and give you a whole audit pack. Like, here's all the data you need, here's how you export it, here's how you give it to your audit. speaker-2: Yeah, I mean I I think Archie's really powerful, just as Mac will be powerful, but I can ⁓ you know, when you actually dive down to what people are doing with our products, you can see Archie really shining. Yeah. These people are going in, diving in, looking for calls. Give me this group of calls from 2017 to 2019 that Michael made with Fidelity. And These guys and gals that use it have to know how to navigate our UI. Yes, we make it we try to make it super user friendly and it's way easier than it just being in some GCP or Azure and them trying to find it that way. It's easier, but can you imagine just typing in what you need and it's right there? So I think and not to dive into the agentics stuff below or in in our next, but I think that's the big thing where you look at workforce management and Mac. Yes, Mac can do that or will be able to do all that. But I think most people want to see a calendar and want to see what what's going on. So I think the UI in that world is still im important. ⁓ where I think there's gonna be some companies out there and some use cases where they're just like, I don't need continuity replay if Mac's hooked up to all the back end anyways. We know exactly what we want. We'll just fire out question or to Archie. Yeah. We'll just fire out questions to Archie and he can give us yeah, give it back to us in great reports. Mm-hmm. So that's why I think it's so powerful. ⁓ which is exciting. speaker-0: And I mean we were I think working our way towards this a couple of years ago when our like our marketing talked about okay, continue replay is creating the foundational data layer that you need for your AI. So this is just taking it that that step further. But you brought up a point with ⁓ c kind of comparing Mac and Archie. I'm curious as to like the approach to building Mac and Archie, how they differ 'cause they're very separate verticals, very separate use cases. Like how how does that approach differ, Bob? speaker-1: Yeah, so it it really comes down to data. So for Archie and and the work he has to do, it's a a lot of big data, like enterprise level data problems. We're talking hundreds of millions, if not billions, of records that Archie has to be able to scan with specific criteria, understand, and then kind of give back to the user, right? So that's a a really a big data problem. ⁓ however, it's it's more restricted, I'll say, right? Because it's always around your customer interactions, right? Conversations that were that are happening. When it comes to Mac He has to understand a larger domain and it's not big data problem. So he needs to know about the people in your center and the stations in your center and time off requests and overtime and what are your policies around those things and how does that map to your payroll and what does that mean for scheduling purposes? So the amount of things that we have to teach Mac about we as an Will Mac as we're building him, how we have to teach him is very different between the two products. speaker-0: Gotcha. Okay, and how do you teach him? speaker-1: That's a great question. ⁓ so when you build Yeah. ⁓ so when you're when you're building or training an agent, ⁓ you're exposing tools to it. So you have to tell it when the user says these types of things, here's how you interpret that intent, and then here's the right tools you have to pull so that you can go and complete that task. Basically what we're doing when a user says something, we want to give a list of tools to the agent and the LLM is basically gonna help it pick what's the right tool to accomplish the goal. speaker-2: Yeah. speaker-0: Yeah. speaker-1: Mm-hmm. speaker-2: And then just like how y you are with Claude, yeah. The more our customers use use it, correct, the better it gets and the more it understands the sound. speaker-1: So that gets into features like memory. Like if you use memory and claude, right, as you have conversations and things are repeated or it it notices something that's really important, it stores that to memory so that it can be reused later, right? So it's kind of self teaching in that way. speaker-0: Right. Okay. Very un speaker-2: And yeah, and just for the audience, you know, you have Mac it is the AI agent for the PSAPs and WFM is the SaaS solution for the PSAPs. Continuity Replay is our SAS platform for the regulated industries and Archie is the AI agent for that. So it's we we have SaaS platforms with AI agents. ⁓ that kinda go along with them. speaker-0: They're kinda separate duos, serve different purposes. ⁓ all right, Steve, I guess I'll I'll pose kind of this wrapping up question with Continuity Replay to you, but is there anything that you wanted to call out on kind of the sales side from the Continuity Replay perspective? I know the guys have been humming with speaker-2: Yeah. So it's it's super exciting right now. Continuity replay plays in a in Archie play in a in a pretty cool world. So most of our customers are gonna be Fortune Thousand companies. ⁓ which is co as Bob said, we need tons of data. So all their customer reactions. So you're dealing with the largest companies in the world. and now we've gotten to a point where, you know, our sales guys that know this inside and out are basically saying, Hey we gotta go big whale hunting. ⁓ so if you look at our pipeline, it's it's some heavy hitters, which is exciting. ⁓ that sales cycles take a little bit longer. but those really make sense for us as to get continuity replay in there. That's our way to get into these large places and then Archie can just expand more and more workflows once we're in. ⁓ so I don't know, I d I didn't I don't want to say anything I shouldn't out on a podcast. Yeah. but we got off to a hot start first quarter with some large notable wins, CBS being one. I think I can say that one. and then now Q Q two, we pumped up for Q three, which we're in right now, which we're waiting on a couple signatures that I don't probably shouldn't say, but ⁓ some large companies and and our pipeline is as big It's way bigger than it's ever been. So that's exciting. that's kudos to the sales guys, ⁓ to Chad, the new sales VP, kind of getting the the new processes going. And then obviously marketing and our brand awareness just getting better and better. speaker-0: Yeah, I think we have to shout out the partner perspective as well. speaker-2: program. Yeah. We get a lot. So these companies we sell into are are large, difficult to work direct. Sometimes just getting it MSA'd that gives you the ability to sell into them could take a year or two. So we work through some great partners, Converge One, Five Nine, WWT. And W and these partners allow us to get into these places a little quicker. ⁓ and kind of expand our sales team. So good call out there. And Jim Giblin For Will Mac who kinda heads that up for us. Jimbo. speaker-0: Jim. Jimbo. all right. ⁓ I guess as we wrap up kind of this this section, Bob, is there anything else you wanna share on the vein of continuity replay or Archie before we wrap up? speaker-1: my excitement for Archie. I'm just super I guess. ⁓ I don't know. It's just it's my mind is blown at the strides that Sammy's been able to take and just what the next few months looks for for us. ⁓ I don't know. speaker-0: I think too I have to I have to call out the UI adjustments. I I'm pretty sure Anthony pioneered that one, Anthony Hartwig. that was just wild when we have our biweekly continuity roadmap meetings. And we just came in one day and Anthony was like, Yeah, I Claude helped me and I basically worked through an entire revised UI. ⁓ and that was pretty powerful and that usability has become like a very consistent theme on the marketing team. Grace is like nit pick in on that. She's like, how do we make everything like this? All our marketing efforts, all our products, that sort of stuff. ⁓ so I gotta shout that out too. That's very exciting. ⁓ all right, wonderful. Let's move into kind of our final, I guess, theme of today's conversation. So ⁓ boy. And then ⁓ so beyond the the products that we're very close to and love to talk about, we've been embracing Agentic AI as a company. ⁓ we gotta practice what we preach. so I was looking back in my emails. Steve originally rolled out Claude or the use of anthropic ⁓ March third. So I guess ⁓ kind of both you have a generic question, but where are you seeing it show up most in how the company operates day to day, besides product and where it's showing up with Mac and Archie? speaker-1: So I can or cannot talk about my team. speaker-0: You can. Yeah. A hundred. A hundred percent. speaker-1: You know, so we started the dev team started using Claud back in November. So a little bit of how to the whole like company rollout. Yeah. And we were just immediately like, holy cow, this is different and it's special and it's gonna be a real force multiplier for the team. we are using it very agentic, so you know, we ⁓ mul oftentimes have multiple quad sessions running. So we're working on two, three, four initiatives for like code development at the same time. ⁓ which is just awesome. So that's where we're really getting the ⁓ Efficiencies and the speed the velocity of each. speaker-2: Person has two to three or four years. speaker-1: Correct. ⁓ Anthony is on a whole nother level. He has six to eight. I feel like I I just can't even look at his computer screen when I walk by. he he managed it very well. ⁓ where we're going though is more autonomy. So one of the things that I talk about a lot with my team is to stop thinking about ourselves as software engineers and more like product engineers or product managers, right? So what we want to do is instead of spending our days at the office prompting Claude to write code. We're spending our days at the office prompting Claude to prep tickets in JIRA. And then overnight, Claude is picking those up. And then when we come in in the morning, we're looking at that. We're reviewing the code and we're testing it and then moving on to the next thing. ⁓ that's where we want to get to. I don't know how quickly we'll get there. There are some tools available to us through like our code hosting platform, for example, that we're looking into that we think are gonna be the key to allow us to get there. speaker-2: I know I think it was Anthony who d was showing me the the podcast he listened to recently with anthropic and someone else. Warp. Warp, yep. But the anthropic guy was basically like, Yeah, we don't prompt for coding at all anymore. We make sure we're we're helping him out with the loops and the loops will prompt. And so th things are are flying on the code side. speaker-0: Instead of forty hours a week, it's a twenty four seven type of thing. Yeah. Correct. Okay. Yeah, that's pretty powerful. And so yeah, d I mean, Bob, I have to say your team has definitely like pioneered the rollout ⁓ at the company and ⁓ and they help everybody. And they help everyone. Yes, always resourceful and ⁓ making sure to to offer a helping hand when in need. And I think too, like we started our weekly now AI kind of working sessions as as a company ⁓ that usually you lead or someone on your team leads. So do you want to talk a little bit about that too and kind of the rollout of how clouds go speaker-1: ⁓ so we certainly recognized for some folks, Claude was brand new and this whole workflow is brand new. So we just started doing every week. We'll get together. I would say it was a little awkward to get started because it's easy for me as a dev team to throw all this technical stuff right in your face. But how do we translate that to a marketing person or a project manager in the work that they have to do? So ⁓ we just kind of took it slow and just started building things together, very hands-on, watch me while I work, that type of thing. And then we shifted to All right, well, bring me use cases to this meeting, right? And let's all talk about them. What's a problem you want to solve? And we're trying to think more like not just me as an individual, but like what would be a good problem to solve for the company as a whole, because that's gonna be more ⁓ it's gonna translate better to the group as a whole. or recently you vibe coded something and I put you on the response and said, Hey, on Friday, can you talk about how you came up with this, how you implemented it, and some of the problems. I think those conversations really were starting to get a lot of traction. We're definitely seeing it pay off. speaker-0: Yeah. ⁓ yeah. I mean when I was working on it that morning and I broke it, but I had an earlier version of it. So Bob was like, ⁓ yeah, ⁓ like show show ⁓ what you're doing. So I showed everyone and we probab we had at least representation on every team. We probably had twenty of our a little over fifty percent of our company there. Neil chimed in, Nick Brewster chimed in, they were like asking probing questions and they're like, Have you thought about it this way? So it offered some a really unique perspective on something that I like was very specific to marketing. ⁓ so it is really powerful and it cre kind of creates like a a seminar of of sorts or workshop where people can actually sh come and show it. And I mean, I know you've put my Grace Basil on the spot multiple times and had her show it and She's b our freaking AI guru on the the marketing team too. So it's it's really fun to kind of see how those conversations are continuing to build. Steve, how's it looked from your perspective? speaker-2: Yeah, I think you know, when we rolled it out, it was like experiment. Just start using this. it's I don't want to say easy but natural for Bob's team to do. I would say Claude and OpenAI, the first industries they attacked, not attacked, but came out after were software engineers. So it makes perfect sense. ⁓ but what I had I'd been screwing around with open AI, Grok perplexity ⁓ before I jumped into Claude. And it was more just experimentation. Experiments. Screw around with it. It's amazing what these what what it can do and what it can help you for. So I knew there'd be a ton of use cases if we just kind of rolled it out and did it. So I think it's been great. I think everyone embraces it. I'm surprised at how easy it is to vibe code stuff. That is crazy to me. And there's certain people that have been doing a great job. Obviously you just mentioned grace. ⁓ but I mean any presentations, emails. I I rely on I mean, I was hooked to Claude for a while. Now I basically have all three Claude, Grock, and Chat GP open AI. Yeah. And if it's that important, I'm running it through all three to see what they think. Claude's definitely still the the primary. And with co code and co work, it has a lot of other features where I'm still although I did just build a scheduling guru. in the open AI age new agent that I'm testing out this week that seems to be working pretty well. but yeah I think it's important again I I I've said it on other podcasts. I say it to the company. I think it's important for every employee to use this. If you're not using it, you're it's you know you're not doing anything for yourself or for your company. You're getting behind the eight ball. I know, you know, we've been we've been interviewing a lot of entry level people or people coming out of college and you're seeing them kind of be like, so are we supposed to do this? Are we not? Our professors, yada yada, yada. And that blows my mind. ⁓ is this isn't ⁓ in the real world it's not about how you get the information, it's about getting the information and getting it done as quick as you freaking can. So ⁓ it's an unbelievable tool and I think we're doing a great job with it. We'll keep pushing it. I do I say this to Bob and I say this to the company as well. AI for our customers come before AI for Will Mac. So we always make sure we're doing what's right for our products before we do internal stuff. speaker-0: For sure. Yeah. And I mean, I know I I always hear this is the fact that the PR around AI is very poor in terms of like the public market and how people view it. It is interesting. ⁓ and I don't know, I l I know I was cautious of it, but like Grace, perfect example. Came out of college, was an intern technology college. I ⁓ yeah, in a technology college, and basically told me that she didn't trust AI and didn't want to use it. And I just kind of told her, Hey, use it for what you're comfortable. For but like what what you're comfortable using it with right now and just keep asking it. And she says to me the other day, she was like, I'm like, I don't even know how I would get my job done without Claude right now. Like the stuff that I'm thinking and that I'm enabling myself to do was not even in like in the ⁓ long distance like from where I was looking before. So I just encourage people to try it to there's free versions of most of those out there just to ask it questions. Like I've used it before to learn how to start my lawnmower when I didn't have anyone to ask and confirm the direction. speaker-2: I would not that we are recommending anyone do anything, but I would recommend do a paid version 'cause the free version you could get some wild policy nations and your data could get out in the real world. So just pay for cheap there's cheap version. One of the big things we'll p we can probably talk about on the next one because we're still getting around to it is this AI brain. That's basically connecting all our systems into Claude and being able to speaker-0: There is cheaper. speaker-2: ask Claude questions that they can go to our Salesforce, our ticketing system, our project managing system, our QuickBooks, and it brings it all together. ⁓ larger companies have been using companies like Snowflake to help bring that together. A company of our size, we could probably just do it with Claude. And so I mean that's awesome. Sometimes when issues pop up, it could take forty eight hours for you to get together with the entire team and get the whole story before you have to talk to the customer or before you have to do what you do. This way you literally just ask ask the AI Brian a question and spits out everything. So there's a lot more we can do that I think we'll be doing relatively quick in the in the next few months. speaker-0: Well I think in that same vein too, ⁓ transparency. Like we had our half year company wide meeting and we always have like our town hall esque slides where we talk about where we're do how we're doing revenue wise, what's closing, what dev team is doing. Like imagine if you're an employee and you can just prompt, Hey, how are we doing? Right. Like that that tells us a lot and that's something that my team has brought up to say like, ⁓ I didn't even know that the sales was progressing on this. That would typically come up in a meeting and they don't know who to ask and I'm like, You could just ask Claude and it could give you the whole lowdown of how we're doing. ⁓ so Bob, I guess ⁓ as we start to to wrap up this quarter three behind the build episode, is there anything else you wanna share kind of on that front and as to how we're using AI within Wilmac? speaker-1: ⁓ I think everyone's crushing it. ⁓ you know, people come to me a lot with questions. I think w what I've loved about our rollout, I haven't really had to like take a ton of time out of my day to day to go help people. I'm able to just give a quick hit. And you know what my answer usually is? Ask Claude. Right? Like when you're stuck, ask Claude, and I'll help you prompt with like what exactly should you be asking Claude to get out of this situation. ⁓ but it's been pretty hands off for me. ⁓ which has allowed me to Steve's point stay focused on the product dev that's really important to us. Yeah. ⁓ so speaker-2: I mean I think that's the what most of the time people are going to Bob and his team is like when they're trying to vibe code something or get cult work to work. Yeah. And you can literally just chat with Claude and it'll help you out. I do I think we've talked about this. It's going a li it's been, I don't know, the new opus I think goes a little gives you too much information. It's gotta tighten up a bit. ⁓ but overall, just chat with Claude Chat and it'll tell you what to do with Claude Code or Yeah. Or cowork. It's it's amazing. Or do open AI for K Claw or whatever you have to do. It'll tell you exactly how to do it. Or at least tr help you troubleshoot it. speaker-0: A hundred percent. Yeah, troubleshooting is a big part of it for sure. ⁓ yeah, so we just encourage our listeners to ask questions, be curious, dig into the a jump into it. That's for sure. ⁓ all right, wonderful. Steve, do you ⁓ I guess I'll pose this last question to you, or what are you most excited about heading into the second half of this fiscal year? speaker-2: If you're in college, learn it. A lot. I think Archie and Mac are obviously super exciting. Like when you see demos of this, it's like, whoa, we've came a long way and people are gonna love this. ⁓ but I think overall, you know, we're so deep into it we don't realize what we've already accomplished with WFM getting a product from literally a thought in in your head or at the customer's head that was helping us to a paid customer is a is a monumental step. So and the pilots that we have going, again are are the big nine one one centers. To be able to lock the three to five of those down, I think the momentum that we will create in the next six months is huge. So pretty pumped, I would say, you know, everything's awesome, but pretty pumped up about where WFM has gotten so quickly. speaker-0: Yeah. ⁓ Bob, I think we've done, I don't know, four or five of these behind the build episodes now. And to every time when I reflect on like what we talked about in the last one and how much things have vastly changed since the over the three months, I'm like every time I'm like more, even more has changed. It's like crazy to reflect back. So thanks for always coming on and joining us and staying ahead of the game so you can you can speak to those changes. ⁓ all right. Thank you to Bob for joining us on ⁓ quarter three behind the build episode. We it's always a pleasure. for our listeners, if you'd like to stay connected with Wilmac and follow along with Wilmac Continue WFM and Mac, or Wilmac Continue Replay and Archie or both, or you just like the vibes that we're emitting from this podcast and want to see more of us, you can follow Wilmac on LinkedIn. You can visit WilmacTech.com for more resources. subscribe to our monthly product newsletters. And you can lastly tune tune into the Will MacWire wherever you listen to podcasts. So thank you for joining us for another episode. And that's it. speaker-2: Awesome. Thank you. speaker-0: Thank you for listening to the Wilmac Wire. If you liked today's conversation, make sure to subscribe wherever you get your podcasts. Or watch full episodes on our YouTube channel. You can also learn more about Wilmac Technologies and our work in customer interaction technology at Wilmactech.com. We'll be back soon with more conversations with the drivers, those shaping our industry, and those building right here in western New York. See you next time.