Your AI Sounds Like Everyone Else's: Here's How to Fix It

RocketDocs: Welcome to Ground Control, the podcast navigating the high-velocity frontier of enterprise AI proposals and governance. Sponsored by Rocket Docs and Luma. I'm your host, Lexi Hotchkin and today we're strapping in with Perry Robinson and Brian Jenkins to talk to you about how to keep your signal strong and your voice distinctly yours as AI reshapes the entire proposal and RFP landscape. Let's get to work. Lexi Hotchkin: Okay, so the last time we recorded, we talked about procurement in our space moving at mach speed one. In the weeks since we recorded, has anything happened in the industry that's proved your point or surprised you? Perry Robinson: Yeah, well, I guess it hasn't been a matter of being that surprised. ⁓ I should been able to predict the potential outcome, but ⁓ yeah, all the geopolitical events are are not all solved. ⁓ you know, there's still questions about whether the straightover moves is gonna be open. people who were traveling to Dubai or Europe are still wondering whether that's safe to fly there, whether there's gonna be enough jet fuel to get there. ⁓ and and then there's questions today, you know. Bryan's located in South Florida, right? And we we now know that Raul Castro's been charged and there's gonna be issues with Cuba as well. So is not a political show at all, right? I just bring these issues up because When you get places like MIT and its Sloan School of Management, or Oxford and its College of Supply Chain, they're analyzing these geopolitical events and how they impact buying behaviors. And the two things that ⁓ we've definitely seen more of is that one, there is a increase in buying behavior because people are having to think about new ways to source new technologies, new pro products. ⁓ ⁓ new goods. And so whether you're you're trying to figure out how to get something to the right place or figuring out where to buy it from, you're having to go and go to market, right? More RFPs, more bids, more tender opportunities. the same ⁓ side, right, you're in existing, if you have an existing relationship ⁓ as a ⁓ as a as a customer in a vendor relationship, you are having to answer more questionnaires about your capability to continue to meet all of the requirements that you were brought in for before. And oddly enough, it doesn't come down to just fit physical goods. Companies are also seeing that there are more questions and challenges to be able to substantiate your capability inside of a pretty challenging economic environment to continue operating. So there's tests for resiliency. ⁓ there's tests to see ⁓ you know the questionnaires being issued to understand is your supply chain going to create additional risk are you already ahead of the curve? Right. But whatever it is, more questionnaires, more questionnaires, more questionnaires. Lexi Hotchkin: Yeah. That's a great answer. That is very true. ⁓ what we're seeing in our space. It's insane. you made the ice cream analogy talking about AI last time, Perry. Vanilla responses. What truly different differentiated AI-assisted RFPs look like in practice? ⁓ Perry Robinson: Mm. Yeah, I've got some bad analogies from time to time. ⁓ but I'll keep on I'll keep on rolling with this one. The reason I thought about it is people talk about things being like plain vanilla, right? And the one thing you don't want to be in a competitive space is plain vanilla. And a lot of times I think people will forget that when it's an RFP or a proposal process, you've been invited to the opportunity to sell to an interested, prospective buyer of your products or services. Lexi Hotchkin: That's a good one. Bryan Jenkins: A good one. Lexi Hotchkin: Yeah. Perry Robinson: Right. And so, yes, is it is it a lot of work? ⁓ you know, yes. Is it the type of work you enjoy spending a bunch of time on answering one question after the next? no, probably not. Right. And and so then people start looking at opportunities to use different types of tooling. And so when it comes into the use of AI technologies, one of the big challenges that people are really thinking about is how do I maintain my company voice? ⁓ Right. The way that we speak as an individual sales representative or as as a collective group, how do we represent our brand, our authenticity, the thing that makes us competitive? And and that's where that plain vanilla comparison you know comes in, is that too often, right, if we go into a claude or a Chat GPT and we produce something, most of us can see which one it came from. You know, Claude produces those lovely little carrots that are blue, right? And Chat GPT loves those little green dots. but they also have writing styles. So even if you delete those or you change the part where it puts in different types of formatting, the voice of AI that you're using can become that's the same for you as it is for your competitor if they're using the same tool. So we've been ⁓ asked a about. You know, well, what else are the alternatives? And of course, the alternatives are when you're taking AI technologies, but you're using them with information that's yours so that when they're producing these types of answers, they're producing it knowing how your organization speaks. So it's on brand, it's on point, it's spoken in your style, it remains competitive. It is not plain vanilla. Right. You've got double Dutch peanut butter super chip chocolate hazelnut sprinkles. Right. Super distinctive represents everything that's you and hopefully everything your customer wants to hear. Lexi Hotchkin: That's a great answer. And I would one hundred percent eat that ice cream. Perry Robinson: You did. Lexi Hotchkin: And you both talked about last time finding your voice before going to AI. For teams that have kind of gone AI first and lost that muscle memory, is it recoverable? How do you rebuild it? Bryan Jenkins: It's recoverable. I think it's one of those things where it's so new and everybody's trying it out and it's it makes you feel so productive really quickly, which is nice. Then when you start digging into it a little bit, you start realizing you you might rely on Lexi Hotchkin: Mm-hmm. Perry Robinson: Mm-hmm. Bryan Jenkins: And then I think people realize that pretty quick and hopefully they're starting to realize it now as you're starting to recognize everybody's using it. so I don't definitely don't think we're too far down the path to where people are gonna start to push back. don't know you s do you guys see that all the kids were booing AI technologies this week? Did you see that? Yeah. I think they're starting to get a sentiment around like AI replacing everything. ⁓ And ⁓ I definitely think your voice or the voices that people have are gonna start to Perry Robinson: Yeah. Yeah. Bryan Jenkins: hopefully make a comeback. Because it's been a lot of I don't know Perry if you agree, but I the stuff that I've seen examples wise, ⁓ just written in general from emails and everything else, just starting to feel like you could tell when stuff's coming through there. It almost feels like the same person's emailing me, like a robot or whatever else in multiple different facets of my life. So Lexi Hotchkin: Yeah. Perry Robinson: Yeah. Yeah, the email inbox is probably one of the biggest ways that we we all end up seeing it, right? So it's the ⁓ you know, classic example of ⁓ ⁓ you know, you're receiving an email that's been written by AI to make it look like it's created just for you based on some information. But seeing some pretty interesting tricks too. You know, people like trying identify, how do I know when it's it's that when it's getting really good? Because sometimes it does. It gets really good, gets really detailed about something that you post it on LinkedIn or Podcast that you were listening to or on. interesting one that I saw was somebody said that on their LinkedIn profile, they added a little line that says that if you want to sell me something, make sure you include your best banana bread recipe. Right. ⁓ And and that person now, like because the bots are going and scraping their LinkedIn profile before reaching out to them. They see that part, they think that it's actually a real thing. And so the last thing is that, yeah. And and not to forget, here is my banana bread recipe. Right. And so now they can see, like, okay, I know when it's AI. but I think Bryan, it's it's it's like there's there's some sort of a a bit of a pushback that we're seeing. ⁓ and at the same time, I'm also seeing that there's also Bryan Jenkins: Even. Perry Robinson: massive amounts of continued adoption. So I hear people at executive levels that are being measured based on token utilization. Right. So it's like, are you using all the tokens that we've set as a minimum token utilization requirement for your business? Bryan Jenkins: Yeah. And that's that's what I love about our approach. I think it's enough AI to help you without replacing your voice or who you are, which is I I really like that approach. I think more people are gonna adopt that over time. Lexi Hotchkin: Yeah, I love that. And kind of going off of that as well, we touched on retrieval augmented generation pretty briefly last time. For listeners who maybe aren't as technical, what is one thing they need to understand about their how their content library actually understands and feeds AI responses? Bryan Jenkins: You wanna take that one or? Perry Robinson: Yeah, I'll start it off. ⁓ So yeah, retrieval augmented you know, shorthand rag, right? ⁓ rag-based technology is a way to ⁓ really get to information that you have inside your own system. It's the most ⁓ simple way think about it. ⁓ and and that's versus having it to where the AI is going out to the internet and finding anything they can find, like you would on a Google search. So for me, if if I'm gonna simplify the concept ⁓ to the most basic level, right? ⁓ ⁓ there's either the library books, sorry, the books that I have in my in my own office space, right? And if I'm saying I just want to source information from those, ⁓ then in this case, the AI is doing the equivalent of just getting access to that versus having access to like the entire New York City public library system with everything that it has on its shelf, which is the equivalent to the internet. Lexi Hotchkin: Yeah, that's great. Bryan Jenkins: Yeah, and I'll I'll add just back to your voice, it's nice because it's finding your voice right before ⁓ and also allowing you to pick different topics, right? So a lot of times AI, and Perry, you're pricing this too, but like when you when you unleash an AI in your whole entire library, if you get asked a certain question and it's relevant to a type of product that you offer or an industry, right? Or maybe even a region, if you just have bulk AI go out, it's gonna find an answer to that question and it might be right, but what if the region's wrong, right? The question's gonna get asked the same way. So Perry Robinson: Yeah. Bryan Jenkins: Rag lets you pick the folder and say, Hey, this is how we handle that strategy in that region. ⁓ and it's gonna pull that in for you. So that that's really helpful too. Lexi Hotchkin: Yeah. Perry Robinson: Yeah, we've all seen that in the in the spelling ⁓ happens between the UK and the US, right? And somebody will pick that out. But interestingly, starting to see it as well in in things where it's a, you know, in the United States, ⁓ the ESG reporting became went from being kind of being a mandatory part of of a lot of businesses to being one that's just not quite as as ⁓ highlighted, we'll call it. But if you're inside of the UK or Europe, it's still important. and so there's a lot more emphasis still on those spaces. oddly enough, what I saw the other day was that there's there's ⁓ you know, people are now starting to say, okay, so we're at the point now where we have to start talking about you know, AI utilization, right? And that lining up with ⁓ you know, goals that were set years ago. ⁓ To target reductions in energy utilization. And so we've got this contrast that I was mentioning before with people who are pressing for more token utilization, with having it to where there's a decrease in the amount of energy consumption that's happening. so if you're sourcing your information for a particular answer, when you're in that competitive process, you definitely want to know whether you're speaking inside of an environment. Where a particular phrasing is going to receive positive be received positively or or negatively. So my US library is definitely going to have a lot of ⁓ details about the performance of my company and how we've made it ⁓ taken advantage of AI opportunities. Inside of Europe, I'm going to speak to my AI capabilities inside of the new AU, EU AI Act regulatory framework. And I'm probably going to want to emphasize how I've intelligently figured out how to utilize AI without increasing the amount of energy consumption inside of my company. Right? Same answer, same question, different, well, no, same question, right? And different answers based on the regional areas. Bryan Jenkins: Yeah. Lexi Hotchkin: Mm-hmm. Yeah, that's great. Kind of pivoting to human in the loop, Perry, you made a very strong case that it's not just a human in the loop, it's the right human. And when we talked to Christina Godfrey Carter, she also mentioned a very similar thing of you have to kind of step up and take that for yourself. And so my question to you is how do you map that inside of organization, decide the right person, make the right decision, who owns that decision? Perry Robinson: But not Yeah. ⁓ well it's got it's gotta be team owned, right? So the human in the loop today, it it it kinda depends in part on on where we want to actually deploy those human level capabilities. I think all AI systems should have a human reviewing ⁓ you know the accuracy of the results. even when it's just automating a process until you know that it's there. But when it's automating responses to things, then you want to have a human checking to see, you know, is the information that's being sourced giving the right result. And and then going back to the competitiveness, right? If we're in a competitive process, we want to move one step further and make sure that you know we still have the competitive voice and the persuasive component built in the way that we need to. So terms of like which people in the organization, you know, it could be down at a at a at a manager level and just thinking about like who on my team, right, is is going to be best to help out with a particular activity. Who's the subject matter expert for that? ⁓ at a company wide level, you know, we've seen that best practices look like really having a cross-functional group. So that you're thinking about those scenarios where You know, on the sales side, I might be thinking just about persuasiveness, but it's also important to maintain, you know, all the regulatory and compliance obligations. So maybe I want to bring in somebody from the compliance team as well and get their perspective on what the best approach is. But Bryan talks to folks, you know, all day long about a lot of this as well. So Bryan, I'm what what ends up coming up when you're speaking with with folks that are really trying to figure out how to how to both deploy AI technology and to have, you know, that human based impact. Bryan Jenkins: Yeah, I mean I talk to teams all the time and they're trying to figure it out, right? Like with AI replacing a lot of the repetitive work, they're trying to figure out like one, how many humans need to be in the loop in the future? What's the ideal amount? And then there's like a a balancing act between productivity of adding the AI, but then also figuring out how to make sure that those responses, like we talked about, actually keep the voice of the company. So what I've seen is like They're in investing or the good, the good teams are investing a lot in people that have ⁓ AI background, that understand ⁓ how to review, that are good writers, good proposal managers. And what they're doing is ⁓ essentially allowing them to be empowered with tools, ⁓ but making sure that that they they're kind of like Natasha on our team is a great example. She keeps people accountable, she makes sure that things are correct on the team, she'd be a great Perry Robinson: Mm-hmm. Bryan Jenkins: And she is our library manager for a reason. And those kind of people I'm seeing are the humans that are are driving this. yeah, and then it kind of comes to subject matter actually after Lexi Hotchkin: Yeah. Bryan Jenkins: that. Lexi Hotchkin: Yeah, that's that's awesome. And so following up on that, what happens when the right subject mat matter expert is also pretty unavailable? How do proposal teams realistically solve that? Perry Robinson: matter experts have perpetually been unavailable ⁓ ⁓ team managers, they would say that's the it's the constant chase down work. So look, for anybody on on that RFP team, it's critical that that I think they understand two things. One, it's the SMEs are not trying to avoid them because they don't like them, right? It's it's they're trying to get other things done, and usually they're not actually given resources that account for the work that they have to do on helping out with RFPs. So ⁓ Lexi Hotchkin: Mm-hmm. Perry Robinson: You know, the process that in involves them inside of the RFP process needs to be one that's easy for them as well. And then it makes it easier for the people on the RFP man RFP team side. Because they they if if you have it to where the SME is able to comfortably get the work done, right, then there's not as much follow-up work for you to do. There is follow-up work to do, and you do have to communicate with them, right? So the second part is how can you do that in a way That's systematized, right? And that takes advantage of the capability to not have to search through your email sent files to see when the last time you actually asked the person inside of, I don't know, the engineering department whether the specifications that were listed are the correct current ones. So we always recommend whatever you're doing, you're using a system, right? The system's got the capability both to store the information and keep it up to date, keep it signed off by the SMEs, but when you need to go back out to them, that you've got an easy way to facilitate communication and manage those communications as well. So you always want to look for a strong dashboard capability. That's able to, at the touch of a button, shoot off an email, right? Also provides in-app notifications for the SME to be able to see when they're inside the system that there's some things for them to approve, right? And then, you know, you know, ⁓ you want to look to the capability to have it to where other bulk capabilities to be able to send out like all the different SMEs inside of one RFP, all the email notifications about a change or a new question that's popped up at the same time. but it comes down to you know single pane of glass. ⁓ Right? One place where you can go to manage everything. Lexi Hotchkin: So what I'm hearing is more resources for SMEs twenty twenty six. Perry Robinson: ⁓ more resources. Yeah. I don't think they're gonna get ⁓ in twenty twenty six with the way things are going. So work yeah, work more effectively, more efficiently. Lexi Hotchkin: Yeah, that's great. Which it actually kind of flows into our next question really well. If the library b ⁓ beats the algorithm, who's responsible for building and maintaining that? Perry Robinson: Bryan, this one's all yours. Bryan Jenkins: I feel like a lot of people ask, right? Like it's between you've got IT, you've got legal, but I think I think proposal teams are really well equipped to do that. ⁓ they've been doing it for a long time and in the age of AI I've been saying this for for a while now. I think their skill set is uniquely qualified to creating very curated and well put together trusted information. So yeah, proposal teams. Perry Robinson: I think the proposal teams, Bryan, have you talked to a lot of folks about whether they're actually going down the path of having a content manager that specifically got the role of of keeping information up to date? Bryan Jenkins: Yeah. It's funny that you say that 'cause I actually had like a it was I don't know if it's called a chief content officer or something like that. It was ⁓ they're ⁓ are becoming those kind of roles and they are joining the calls most recently. So yeah, it's it's neat to see. I think it's stuff's changing so quickly. So there's new roles titles ⁓ the industrial revolution, right? There's a bunch of new roles that nobody saw coming and now a bunch of new AI roles that are coming, ⁓ and stuff that didn't used to be necessarily Perry Robinson: Okay. Mm-hmm. Yeah. Bryan Jenkins: role worthy are now very important critical roles. Lexi Hotchkin: Yeah. Perry Robinson: Heard of that as transition roles as well for people where, you know, people are thinking about where they've got valuable team members and they're trying to figure out, you know, where do you put them? ⁓ and then I've also heard for new hiring practices on the sales side, you know, I heard the marketing side. One of the ⁓ chief marketing officer was talking about how, you know, in the past she would have been pushing for additional positions to open up for people to be content writers. Right. And and then what she said what's ⁓ really changed is that today she wants to have content strategist. Right. And so you need to have somebody who can help to think about what needs to be written, because of course generative AI is able to do a lot of the writing work. and so they've got a capability to do strong persuasive writing. So now she needs the content strategist, who of course has the capability to be the human in the loop and to revise and reframe, you know, ⁓ pieces as needed. Lexi Hotchkin: Awesome. I love hearing about all the different jobs that are opening up. That's so insane. How quickly it's all changing. ⁓ please. Perry Robinson: I gotta add one more in there then just because it got shared with me. So had another CEO this morning share with me little interesting study between Meta ⁓ IKEA, right? And so in each of the circumstances, about 8,500 jobs ⁓ were deplaced displaced by AI technology, ⁓ Or AI being you know play the investment's happening. ⁓ but ⁓ Lexi Hotchkin: Yeah. Perry Robinson: Big difference between what they each chose to do. So on the IKEA side, it was customer support folks that were actually working inside of their call center and were helping out with different types of things, probably helping people to understand those crazy IKEA instructions. But AI was able to provide a solution that's able to do like I think 47% ⁓ the work or something like that. And 47% ⁓ seem like that much, but that represented those 8,400 jobs. IKEA chose to retrain those individuals over about two or three month period of time and to give them the capability to start working inside of different areas, which is interior home design with IKEA using AI technology to help actually do it. So these folks didn't have to be designers themselves. They just had to be trained on how to deploy that technology. And apparently they created a whole additional $1.4 billion revenue stream. In this new AI-driven design and decorating capabilities that are offered via their website. Meta, on the other hand, I think, as we've all seen, has ⁓ gone down the path of just going with a straight headcount reduction. ⁓ and so they're they're not taking those roles and retraining and deploying them into different functions. ⁓ the argument that was being made by the person who's reporting on this is that it was a real missed opportunity. So yeah, AI is having strong impact on that side. What we see inside the RFP space is that when people are really thinking about it, they're thinking about what do I actually need to have skill set-wise? And I think the ones that are really thinking about it the most intelligently are the ones that are thinking about how to take existing current team members. Bryan Jenkins: Yeah. Perry Robinson: That are great team members they have, and to help them adapt with new skill sets to be able to deploy them in a different way. So the content writer becomes a content strategist, right? the person who was previously doing a lot of cut and pay copy and paste from something into an RFP response is now the person who's looking for that competitive edge or difference and is doing some research work and then helping to deploy a much more effective proposal or RFP response. ⁓ Lexi Hotchkin: That's awesome. I love the way it's enabled so many of our clients to work faster, work smarter, work better. Bryan Jenkins: I think that brought up a really good point, Perry, but I do think proposal teams and people listening like should should be thinking about like what kind of vision that they have. 'cause I do think it's almost like the art used to be additive, like you're painting and you're taking paint and you're trying to draw it. Now it's almost like you're sculpting. Like you've got access to every bit of information. So how do you remove stuff to get to where you need it to be? How do you filter out the noise? But you need that vision of what that final product needs to look like 'cause AI is gonna give you whatever it thinks is best and if you're lazy then Lexi Hotchkin: Yeah. Bryan Jenkins: It's gonna give you the same thing as every other competitor, but if you've got a vision you could probably use it in a way that will make you stand out. Lexi Hotchkin: Yeah, that's awesome. Perry Robinson: Sure. Lexi Hotchkin: Switching gears a little bit, we completed our third podcast episode, which means we beat ninety percent of podcasts out there. Look at us go. In the three episodes, is there a topic in the RFP or procurement space that you think nobody is talking about yet, but should be? Perry Robinson: Yeah. Bryan Jenkins: Top ninety percent? Perry Robinson: Avec. Yeah, I've got I've got one. Yeah. It's ⁓ so I I think one is is truly on top of mind for a lot of folks. So we hear more and more about people that are working inside of systems that are AI, you know, systems for actually like creating those RFPs and getting them out the door. yeah, it's it's a it's an interesting thing when people start talking about like like Lexi Hotchkin: Hit us with it. Perry Robinson: How is it that we keep up with these? And more importantly, ⁓ how do we actually write things for machines to grade and evaluate instead of people? ⁓ and I do think those conversations are happening, but it's amazing how structured or organized they are across the industry. Like what What do you need to do to write an effective RFP when you don't know whether your audience is going to be a person or an AI machine? And the answer, of course, is that you have to write it for both. But how do you do it? That's that's the that's the part. Is that it's easy to say what the the outcome needs to be. It's hard to say how to get there. Lexi Hotchkin: Yeah. Bryan Jenkins: That's a good point, Perry. It's like like websites when they first came out, originally they were just thinking about the person reading the webpage, which was fine, like in the nineties. Then all of a sudden Google was like, We're gonna make it easy to search and now it's like, How do you balance the robot scanning it versus the person reading it? 'Cause you wanna add a bunch of keywords for no reason. It sounds off. That's a similar problem. Perry Robinson: Mm-hmm. Yeah. Mm-hmm. Lexi Hotchkin: Yep. I was gonna say, now we have to optimize for AEO and that's a whole another can of worms. Perry Robinson: SEO becomes AEO, yeah. Bryan Jenkins: Yeah. Mine I think like ⁓ something that I'm noticing that I think people are maybe starting to talk about but haven't really fully I guess considered is ⁓ like how much to Perry's point the customers on the other side are gonna be utilizing AI to review it, but also are they gonna be asking questions that are more AI specific, like did you use AI to complete this response? Or you're not allowed to to use AI to complete this response, like a public AI. and then how often are those Perry Robinson: Mm-hmm. Mm-hmm. Bryan Jenkins: people that are issuing the business, because they tend to drive right what the rules are because they're the ones giving the money out. ⁓ are they gonna lock down the ability to to use like an open model because they don't want obviously like a large company doesn't want their RFP to get ingested into a a public model. So if you're relying on that kind of stuff, I think thinking about it now it might be okay, but in the future might be something that if your whole process is aligned on on using one of those larger tools ⁓ could be really bad overnight for you. So keep that in mind. Lexi Hotchkin: Mm-hmm. Perry Robinson: It's I think it's funny, Bryan, you say that obviously, right? They don't want to have information inside of a public model. But I think, I think, yeah, the other part that's the elephant in the room is that you know there's been a lot of direction, just go ahead and adopt AI technologies, which has resulted in a lot of shadow AI utilization, right? It's that resulted in a lot of people being pretty loosey-goosey with where information's going. ⁓ and so you know, the elephant in the room is this. Bryan Jenkins: ⁓ Yeah. Perry Robinson: you know, this is kind of like a musical chairs situation. At some point the music's gonna stop and you know, we're gonna wondering now, like where are people gonna find that they're the one without the chair, which, you know, could be the scenario where ⁓ you you know, there's just a a point where we rely so much on AI and and then there's some major major issue, like, you know, some of the Microsoft systems that have been used for running airports and then it stops working and flights get canceled. ⁓ What happens when there's a major power outage inside of a region and all of a sudden critical systems that are AI powered are no longer working or accessible? What happens when there is a data breach that's a result of it and you longer have, you know, there there's a lot of parts that I think have been playing into ⁓ the current approach of like it's good enough. Like I should it's I should be fine. I I and to your point. People though, companies that are issuing these, they're getting smarter and they're thinking about this from the perspective of, wait a second. I actually have to think about whether a company I'm sending an RFP to is going to maintain their confidentiality obligations about the RFP process. And it's not whether they're going to openly share that when they're walking through the airport and have somebody, you know, hear something they shouldn't. It's is it going into a system? And then is that system going to be able to, you know, start identifying that pattern? You and I have talked about the. Lexi Hotchkin: Yeah. Perry Robinson: The big problem, which is when it's a couple of large publicly traded organizations, right? And one of them's about to have a major contract canceled. What happens if that information gets out and people start actively trading shares based on the assumption that there's going to be a a change in the price of that stock because they know that they're going to lose the biggest customer that they have? That's that's a pretty big deal. Bryan Jenkins: Yeah. Lexi Hotchkin: Mm-hmm. Bryan Jenkins: Yeah, I mean that's stuff I don't think people have really considered yet, fully fleshed out. So see how it goes. Lexi Hotchkin: Yeah. It'll definitely be interesting to see where it continues to grow and then where it continues to shrink. 'Cause you have those areas of ⁓ there's some times where it's ⁓ grows so fast overnight and it's being adopted by everyone ⁓ kind of constantly. And then there's some teams still aren't to use AI in totality. They can't even use private or like our AI, because we have a very secure system, ⁓ even still they use those. So it's gonna be interesting to see ⁓ how those teams some teams fail and some teams continue to grow. Perry Robinson: Mm-hmm. Yeah, and Lexi I say think I think one of the other ones that that comes to mind is is that ⁓ I kept on hearing for a while that people were saying they're like, ⁓ well, we we're just gonna build this because we don't want to spend the money on building out this solution or sorry, buying the solution from somebody else. ⁓ and that build versus buy discussion for a while felt very easy because where development resources, it used to be that it would take them like a year and a half to create something. It goes down to let's call it six or eight weeks? Yeah, that's a huge reduction in the amount of time. ⁓ but if we think about those resources, how many companies are actually continuing to add additional technological resources to staff and support each of those? And if they're able to build something in eight weeks that used to take 18 months, how many additional products will they build that they have to maintain? And are they able to think through all of the different facets of what makes something actually work well, right? And makes it secure and makes it resilient. So ⁓ my favorite one, you know, Bryan and I were on a call with a with a cu you know, prospective customer and and it was a great comment ⁓ about you know sometimes, right, your internal IT building something for you gives you something that's like being half pregnant, right? It's not it's far enough along, but not quite there. Right. And then the odd situation is when it's not getting that, they give you something that's fully completed, but then they're not gonna give you ongoing support afterwards. So When they turn it over to you, it's ready to go and it may even be great. ⁓ but six months down the road when it's no longer relevant, no longer the most advanced capabilities, and you ask for something else, they're already off to the next project. They don't have the resourcing to go back and do anything more than do basic maintenance. And you're left with tool that now is no longer gonna make it to where you're able to do, you know, whatever the best, most current version of of you know that technology is Lexi Hotchkin: Mm-hmm. Bryan Jenkins: It's also hard to go back and say, Hey, can we buy it now? Like, no, we already moved it. What do mean? Like, use that. Lexi Hotchkin: Yep. Perry Robinson: And it did cost money. It cost time. It cost the, you know, like it tossed it cost the compute. I mean, all all whether it's us or any other provider, right? We're baking in typically the price of all of the compute. Into there in addition to the cost of being able to keep the information, you know, keep the system secure, up to date, ⁓ to be able to lead with new product enhancements that keep things at the you know cutting edge. ⁓ and then there's the staffing for support and having somebody who's able to answer questions when they come up. It's another thing people don't think about is that we build it internally. Who is the support team? Right? Who's going to answer the questions? When you get locked out, who are you reaching out to? Do they staff your company's internal? IT team that's building things around the clock so that people who are inside of the UK or inside of Australia, New Zealand are able to reach someone during their working day, or do those people have to figure out a way to get you know resolution? And what we've seen is that they typically are saying, well, we've just got you know, submit an email. Submitting an email is great and fine and waiting 12 hours, but not when it's on a critical system. So if you have a deadline on an RFP, do you really want to wait to see if you can get somebody Lexi Hotchkin: Mm. Perry Robinson: to answer your ticket that you submitted on Friday afternoon on Monday morning after the deadline's already passed? Or do you want to be able to reach out to a company support line, get somebody to help you out, and you still get that Friday, five o'clock timeline met? Lexi Hotchkin: Yeah. Bryan Jenkins: Man, you just you brought up something that I didn't really think about that long term is something that I another thing that I don't think people are thinking about. But companies right now, especially in this economy, are probably trying to save as much as they possibly can and building their own things with these AIs, right? But I wonder how many companies are gonna lose focus of what is actually making their company profitable, like their main product, the main thing that they do well, which I think now is more important than ever. Lexi Hotchkin: That was a perfect answer, yeah. Bryan Jenkins: 'Cause there's it's so easy to say, Hey, let's go replace our entire tech stack. Where can we save money? How can we cut people? Back to the IKEA versus whereas there's other visionary people that are like, How can we reutilize these people to make even more money? How do we use our resources to do the things that we're not doing right now that we could be doing rather than saving money on building a tool that's already exists? That's not that expensive to just buy. Like what is the opportunity cost of getting distracted trying to save money, especially Perry Robinson: Mm. Lexi Hotchkin: Yep. Perry Robinson: Mm-hmm. Bryan Jenkins: when economies I get it right now, you you wanna save as much as you can. But what's the hidden cost of of doing that? Lexi Hotchkin: Mm mm Bryan Jenkins: Yeah. I think long term we're gonna find out because a lot of those companies that are building their own solutions are gonna be like, Wait, we forgot about our own product and now we built a bunch of tools inside, but none of them they're all half pregnant and now it's like we don't have any good tools and we've got one guy left doing all the tools with the AI co pilot thing. Like it's just it's kind of a self fulfilling prophecy there. Just keeps cycling down. Lexi Hotchkin: Yeah. Perry Robinson: Quite. Lexi Hotchkin: Yeah. Yeah. Yep. Perry Robinson: Cloud code's good it's not that good. Lexi Hotchkin: Yeah, exactly. Bryan Jenkins: Yeah. Yeah, 'cause it makes you think you can fire the not fire, but get rid of most of your dev team, but then they also need to do more with less and save money on the software and it's just like it seems like a great idea until five years from now when nothing works anymore and you've only got one person. Yeah, not good. Wow. Perry Robinson: Lexi I'm gonna throw one more thing there for us quickly. Lexi Hotchkin: Please. Perry Robinson: Yeah, it's it's a so there's one other interesting part that has been coming to mind lately. ⁓ we call it company knowledge and we talk about it for you know RAGs RAG based utilization. but interestingly when you look at companies like Stack Overflow that have had to go through and think about a way to pivot, because effectively what you know they were is a information exchange platform for folks. it's how do you put knowledge and information to work? Right. And one of the things that we've seen that's really, really an interesting one that I'd love to, you know, Bryan, hear your thoughts, like where you're hearing customers talk about it, is people are starting more and more to realize that, you know, whether you call it garbage in, garbage out, right? Or just saying the better curation of your information, the better results for you're gonna get from your AI. you know, ⁓ are the ways that that ⁓ hearing customers think about how they can actually take. The information they already have and put it to work in new and novel ways to make them more competitive in the RFP and proposal space. Bryan Jenkins: Yeah, absolutely. And and to your point, yeah, the teams that are thinking about that are the ones I think that are winning the most. As far as when I talked about win rates and everything else, it's the companies that have really put an emphasis on saying, Hey, how can I create a library that's up to date and how can I use that in the most novel and unique ways possible? So ⁓ in our case I think it's setting up workflows, making it easy back to to SMEs, right? Making it easy for them to update and then really holding like the team accountable to a process that's easy. feels natural to the workflow that's already there. Like I think that's kind of the key because ⁓ a a lot of the times like you if you try to force people to do things that they're not comfortable with, especially people have been there a long time, but adoption is hard and keeping content up to date is I g I think one of the most important things right now in this climate and this ⁓ the future of what the economy and what companies are gonna be moving forward. Perry Robinson: Mm. Yeah. Mm-hmm. Lexi Hotchkin: I know. For sure. Perry Robinson: On that end, like there's one other part that kind of comes up to it too, is that people have said, Okay, well, there's there's companies like Gleen that are out there that are linking into every single place where you have information stored. ⁓ and then there's a lot of folks that are really talking about, yeah, that's great. That's just kind of like a a different, smaller version of the internet, but how do we get it to where we we keeping information up to date, accurate? We know that it's still sourcing it, even though it's our information, is it still the most current information that should be used? So on that side, like have you have you heard ⁓ any customers talk about best or better practices with regard to figuring out, again, it's it's a you know, the wider data sour data source gives AI more things to choose from and to use information from, but the more curated information source makes it to where the information that's being provided is actually accurate, reliable, and up to date. So How are how are folks that you're talking to handling, you know, kind of that binary path? Bryan Jenkins: the ones that I'm talking to that are handling it well, ⁓ they have the integrations that are pulling the most up to date information, but they definitely have one source of truth that they keep up to date. So they have to have one place ⁓ is ⁓ a human in the loop that approves the content regardless of where it's coming from and it needs to stay up to date in that one spot. And that particular thing, the teams that are doing really well, like I mentioned, have workflows ⁓ making sure that that content stays up to date. ⁓ they also have ⁓ Perry Robinson: Mm-hmm. Bryan Jenkins: Mm, the ability to trust it is another thing as well. Perry Robinson: Yeah.