HITT- AI adoption and compliance strategies for businesses- Feb 18, 2025

The video discusses the immediate relevance of AI technology and innovations, featuring a panel of experts from Telarus who share insights on market impact, data readiness, and security concerns. As AI conversations evolve, advisers are encouraged to guide clients through the complexities of implementation, emphasizing the importance of data cleanliness and accuracy. The rapid growth of AI solutions presents both opportunities and challenges, with a focus on understanding client needs and IT methodologies. The panel also highlights the significance of specialized AI solutions and the necessity for organizations to develop clear AI strategies. Overall, the session underscores the critical role of advisers in navigating the fast-paced AI landscape.

Transcript is auto-generated.

Well, as always, your comments and questions are welcome in the chat window for our live Q&A, which follows today’s high intensity tech training. Today, we take a deep dive into the explosive growth of artificial intelligence and how Telarus advisors can assist their clients in using AI to its full potential.

We’ll show you today how to get better context, greater security, and how to use AI in synergy with clients’ existing data.

Today’s hit presentation is hosted by Telarus VP of Cloud, Koby Phillips. Koby, it’s great to have you back once again. You always fire up the advisors with powerful information and terrific guests, some of which you have again today. Welcome, and I’ll let you introduce today’s additional panelists from Expedient. Hi, everybody.

Hey. Thanks, Doug.

Through the, troubles of air travel, I’m here today. As you guys saw, Dan was originally gonna host, but we’re able to, relieve him of that duty, and, thanks to Delta and some delayed flights, I get to join Brad Reynolds and Tom Cooper from Expedient today to do a deep dive into their AI platform, which is taken off very quickly, right, Brad? I mean, you guys are just seeing a ton of success early, and really what I love about it is how simplistic you make it for customers, a very complex technology, very, methodical way that you guys approach it to make it simple for your for your clients. Can you guys, you know, a, introduce yourselves a little bit, titles, all that good stuff, a little background for the group, and then we’ll dive into what is going on at Expediant and why everybody’s so excited about this AI platform.

Yeah. Thanks, Koby. So Brad Reynolds, cofounder of Expediant, chief strategy officer, SVP of artificial intelligence.

May have another title coming down the pike. But, kind of focus area here is I I built and sold the company, a long time ago back in the telecom days and, came back to to build the AI practice out. And, you know, it’s just been, I think twenty twenty four from our perspective with AI. It was a lot of education and searching and people trying to figure out their strategies.

And it, you know, it definitely feels like and we’re seeing the momentum from the adviser community that twenty twenty five is the year of execution, you know, and funnels and deals and all of that kind of back that up. So that’s that’s kind of what we’re up to. And then I, I recently brought Tom Cooper about six months ago in. So, Tom, maybe you can introduce yourself and and what you’re up to and a little history.

Sure. My name is Tom Cooper. I am heading, AI product here at Expediant.

Brad and I have had multiple ventures, that we worked on together after, he initially sold Expediant.

And now my role is to help build, our AI software solution, not only to meet our direct customer’s need, but more specifically to meet our partners’ needs, to make sure that the solution we have, is easy to implement and solves real problems, for our direct clients and for your direct clients.

Sorry, Connor. I didn’t need to step on you there. So No. No.

Go ahead.

Everybody’s super excited about the possibilities of AI. Right? We have that’s one of the biggest hooks along with, like, some cybersecurity things in in cloud at the VMware platform, changes and everything, which Expediant also does a really good job with. But in particular, on the AI side of things, what have you guys designed that has been so eye catching to customers and partners alike? We’ve seen some very quick wins. Last week, Jason Kaufman actually referenced one that we’ll get into a little bit more in-depth here, momentarily. But what is it that’s happening over at Expediant with this platform that’s that’s so different than some of the other guys that are, going to market?

Brad, do you want me to take that one?

You on the platform.

Sure.

So, we’re approaching AI, and I think everybody wants to say that they’re approaching technology, and simplifying it. Right? But for our perspective, there is a need for security and a need for compliance to be wrapped around AI in order for enterprises today to even step into AI, and that represents a significant challenge of unknown as, you know, enterprises move through the decision making process of, you know, honing on to which AI platform to to make their strategic bet on. So what we do is step that back by providing a platform that actually brings in all of the AI providers but brings them under a, you know, a secure, private, and compliant usage model.

We make it easy to start engaging with AI. And, you know, if you call it a crawl, walk, run approach, that crawling, we really help customers be able to make that step into crawling without worrying about whether or not their data is going to leak or if their usage of AI, you know, is is violating one of the compliance standards that they’re looking at.

And those are two of the major concerns. Right? Tom, and, the ability for companies to start adopting AI or to utilize, it’s almost paralyzing to a degree.

So when you say use all the AI platforms, can you give some examples to the audience like what you’re referring to there?

Sure. ChattGPT four point o, anthropic cloud models are probably the talk right now.

You know, when we talk about Google models or the meta models bringing in each model that has a different flare and something that it’s better for, but definitely at the top echelon of models.

This is separate from the discussions of deep seek and whether or not people wanna trust that. These are bringing in models that you already know and trust the brands, but we’re building the compliance around them and the privacy at the stations around them.

So, Brad, would you say similar to how cloud adoption went, where everybody originally went out and said, we gotta go to Azure, we gotta go to AWS, but as it matured, you started seeing organizations utilize a hybrid approach and utilizing the best of every, you know, breed that’s out there. It sounds like what you guys have built allows for a customer to use all of these public LLMs and, you know, platforms and bring it in and then securely let that resonate within the, the environment for the customer.

What Yeah. Yeah. Yeah. So I mean, maybe think about it this way. It’s a little different from cloud. So when Hyperscale really took off, the things that they did were relatively well understood.

So it’s like, oh, I like Amazon’s tooling or I like Azure’s kind of application set. But the actual instances and services were pretty well understood. With AI, most people don’t understand kind of who’s going to be the leader, what’s the right technology, how do I hook it to my data. So at least for the foreseeable future, having the optionality of trying to use the best tool for the job and letting somebody like us who build a platform figure out how to integrate all these different components into the platform for you so you can just pick the right tool off the shelf when you need it. And you’ve essentially hedged your bets against going all in on one particular provider. Like, for instance, if you go the Copilot route, you have now gone all in on Microsoft and all in on OpenAI.

We don’t know there’s no expert that says, well, that’s that’s the right bet for the next five years. It’s it’s it’s too ambiguous.

So that’s where we look at it. It’s like, hey. We’ll we’ll have a neutral platform.

You can connect to a whole bunch of data sources. You connect to a whole bunch of AI models. As new things come out, we’ll add them as plug ins, but you kind of connect to one platform, and you get the best of the the whole world.

Now one of the major things that we see with customers whenever we start to get into these AI discussions, the one of the first questions we ask, what’s your data look like? Is your data ready for AI and everything like that? Now I know in our previous discussions, Expediant’s not gonna go through and cleanse out the data and everything, but you guys do have a unique way that you’re approaching this as well. So how are you guys helping customers get ready to start consuming and and everything with their data, and making sure that that’s ready to go?

So we’re we’re we’re tackling this from two approaches. I think, as we talk to each company, you know, whether or not there’s been data curation is always a big concern. So the first mechanism that we’ve done is, hey. Let’s take traditional enterprise search and solve this in the most simplest way possible.

You know, let’s give you guys a a new repository where you can bring cure highly curated data into it and automatically have that data vectorized and put, into the chat application. Right? So now you can use your data. Think of HR.

Right? HR is highly static content, maybe changes annually. It’s a limited subset of data, anywhere between fifty to a hundred documents. But the amount of benefit that an HR department gets by taking that data and in one day having that data index where all of their employees can now do things like tell me about my PPO, compare our PPOs, tell me about my four zero one k, how do I schedule a dentist appointment, am I off on president’s day.

Right? All of those questions now become incredibly simple to answer from a technology perspective and incredibly easy to implement.

As we move to the other side where there needs to be a larger swath of data, LLMs are good at unstructured data, but not great at uncurated data because it’s looking at the content of files. So we work with companies to take their most curated folders. You know? You can think of, you know, your IIS documents, your support documents, your training documents, things that have been naturally curated existing where they live, and we can still bring those in and activate those without needing them to copy them. But curation is definitely something that we’re trying to help people handle in a more bite sized approach instead of having to look at AI being on the opposite side of a data lake.

So, essentially, that was that kind of that crawl, walk, run, scenario that you’d mentioned in your opening statement. So whenever you’re seeing this and what’s the engagement look like, guys, whenever you guys are getting into a deal and you’re starting to get into, you know, get some use cases and things like you laid out a couple right there, like, hey. Just make your existing static data more intelligent and useful. I can tell you that everyone in our HR department, given that I just did exactly what you said, I instant messaged, our user messaging platform last week, to, hey.

Can I get a copy of my insurance card? Did you go through the app? Well, the app’s messing up. Cool.

Well, here it is. I I mean, I obviously just ain’t gonna have my own insurance card, but it would have a lot of information there.

Anyways, that, you know, what are some other examples and what’s the engagement look like, Brad, when you’re getting into these deals, with an advisor and walk us through some engagements that you guys have already had some success with.

Yeah. So I’ll let Tom Tom speak to the specific engagements. But, like, the key here, and especially as Tom’s talking about the data part, is how do you simplify the onboarding? Because if you simplify the onboarding, people can connect what’s the ROI to it and then see, like, okay.

I see now how I can get some initial value, and then I can see how it expands. So whether that’s from a client perspective or a trusted adviser perspective, expansion equals more revenue.

So the way that Tom is pretty much responsible for this, that we simplified the onboarding. The onboarding is, hey. We’ll set you up with the service.

We’ll create a whole bunch of folders, like, in SharePoint usually for all your different functional areas of the business, sales, marketing, HR, you know, enablement, all that type of stuff. And that’ll be the starting point. So let’s not get super deep at the beginning, all these potential use cases, and just get caught down a rabbit hole. Let’s just look at some very database workflows that you have and figure out how to automate those.

That’s the crawl. That’s what people are like, hey. I can justify how much this platform costs just by solving the HR workflow or the onboarding of salespeople and enablement workflow or whatever that is. So that’s the context of the conversation initially.

And then kind of more of the walk phase is, okay. And we can also integrate to your ERP, or we can integrate to your Salesforce or to your CX system because oftentimes, just those static repositories don’t show you the whole picture of the company. You might wanna say, hey. I wanna take this out of the CX platform.

I wanna correlate that to Salesforce. I wanna create a QBR about, you know, how this client’s doing. Great. Love that use case.

That is the horrible initial use case because there’s just a lot of dependencies. So this gets you doing something very simple where your whole business starts to understand how to use AI. And then let’s, during that process, start talking. And who is being talked with?

Well, the TAs are being talked with. We’re being talked with on a technical level. But it gives you this kind of future dialogue because in a platform, the pieces that you pay for are the AI utilization and the data hosting. So as you start to connect to more data and you start to use AI more, kind of the recurring revenue benefit to the TAs and to us increase over time.

And, Tom, I don’t know if you wanna talk about any or, Koby.

Some of the other folks or something. Tom, I I definitely wanna hear the examples, and I I know the audience does too, but I do wanna double click on something really quick. If I’m an adviser, what I just heard so we talk about the pressure that AI puts on organizations.

You’ve got a lot of CIOs or VPs trying to figure out how to answer to their other executives and their business on what they’re doing around AI, and what’s a roadmap look like and everything. What I’m hearing from both of you guys is the ability to advise a company that, hey, I can can be a release valve here for that pressure, and we can do so in a very, slow, methodical way that gives you the ability to intake the goodness of AI without having to go do a ton of cleanup and work. That’ll eventually need to get done, but this will allow you to answer to your company of what you’re doing and start to get some motions put in place to give them a little bit of muscle memory or to build that up on how to utilize AI and not have to worry about the security piece, which, Tom, I’ll definitely wanna dive into what you guys are doing specifically and any key talk tracks as the advisors can use to introduce that, and number two, that regular compliance that’s continually moving.

I mean, that’s a motion that is not set, there’s a lot of different company, or a lot of different eyes looking at what changes need to be made as this technology is moving so quickly, and the long and the short of it is, this is a nice, I wouldn’t call it an easy button, right, but it is a nice, easier conversation to introduce to your clients. Tom, what are you seeing whenever you guys are in the deal and taking the approach of this. Can you walk through a couple of those win stories for the audience as well?

Yeah. I I totally can. And we’ll save the the deal that you wanna talk about for after we can go into that, deeper. But you you hit the you hit the head on, on the on the nail.

Right? The the talks that we have start along the lines with every AI talk with we have a fear of GRC, standard governance, and then we have this big these big dreams that we wanna build, and we don’t know how to get between here and there. Right? It’s why the stats out of Gartner say ninety two percent of CIOs say they’re investing in AI this year, but then Salesforce says eleven percent of them have a plan.

Right? That gap is we don’t know how to handle governance, and we don’t know how to baby step into, into getting to what we want to deploy, you know, long term out of AI. For us, it almost is an easy button, and it’s you know, when you look at this as a partner, as a TA who’s who’s going out to move this, the fact that we can take and disarm the concerns about governance by providing secure compliant usage with no fear of vendor lock in puts CIOs who are traditionally difficult to sell in in a position of understanding the technology and being comfortable with the security. And then the baby steps gets all of the rest of the organization a way to move in.

So you mentioned a bunch of them, but HR is a huge low hanging fruit for enterprise search.

Once you start mentioning to customers, you can not only baby step and give everybody access to a chat GPT like client that gives access to every model and is secure and compliant in usage. But on top of that, we can give you low hanging ROI on enterprise search. We’ve been trying to solve as an enterprise, as an as a as a as a company to try to solve enterprise search for fifteen, twenty years, and it’s very it’s proven to be difficult. LLMs have made that so easy right now.

So, you know, sales data, marketing data, when you wanna talk about onboarding data, training data, service manuals, all of these represent generally static content that may grow, but the the traditional content isn’t changing much. And every one of these has a huge impact of ROI on a company if you can knock those out. So we’re basically we are providing an easy button because we’re letting them push the big company pipeline stuff off a bit and find ROI today and find purchase in a secure, compliance system. And it’s really turned into, you know, and it’s gonna sound ridiculous on this call unless you’ve been on a call with us.

One or two call closes, ten day closes on AI on new client engagements. Right? These are amazing Good. Stats.

That actually doesn’t surprise me. You’ve got this this feeding frenzy of people that are looking for for a way to do this. There’s still a ton of paralysis as far as being fear to move. You guys are removing a lot of that fear again and giving them an answer to go back to the company and say, Hey, we’re doing something, right? And then the expansion part is the fun part, seeing how much they utilize the technology and where they go with it from there. What, what is it, you know, in particular, Tom, like if you’re giving a TalkTrack to an advisor and you say, this is how we secure it, this is what we do, could you simplify that message so they can take that to their clients and say, hey, here here’s how they’re making AI easier on the security side?

Yeah. We can talk about it from two angles. First is attestations of non training. So all of the vendors that we work with have a public facing attestation of non training if you’re using their corporate models.

Right? So you’re using it to the pipeline that we provide. So we monitor that and put that in front of our clients so that they know every one of them is SOC two compliant. Every one of them is HIPAA compliant.

Every one of them has public attestations to not train. So you can be confident that our compliance team are monitoring on every one of the vendors those statements. So no matter which model you’re using, you can use your own internal company data because you’re not feared fearful it’s gonna be training on trained on it. That’s the first concern.

So that’s pipeline a. Pipeline b is just an understanding that if you’re put up to a SOC two audit, let’s say, the intent of a SOC two audit is as you move through the audit, show me where the data came from, who had access, and where the data went to. Since all of the enterprise tools that includes Copilot, ChatGPT Pro, Anthropic, Perplexity, all of their clients provide only quantitative data for reporting. They don’t provide qualitative data.

They tell you how many tokens were used. If you’re put up to an audit and the data moves through one of their platforms, today, you can’t show that to an auditor, and the company would fail that SOC two audit. Now I’m sure coal fired and other organizations are working on fixing this. We solve this by putting enterprise monitoring and auditing as a umbrella over all of the tools.

So, basically, this enterprise observability layer just allows a compliance person to understand at any point for the next seven years, They can look up who, when, what model, what was actually requested, and what the actual response was so that they can be subjected to an audit on any data that moved through one of the AI, tools in our platform and pass that audit. It’s that compliance check mark. I know that’s way bigger than a simple statement, but that compliant check mark that really CIOs go great. This is this is the thing we’ve been looking for.

We don’t have to build this ourselves or buy a technology solution to secure, you know, kind of the back end of our entire organization to do this.

So if I were to summarize it, right, and that was a great amount of information, essentially, the way that you guys are tagging with compliance is putting an umbrella that tracks the data through whatever platform it it runs through for up to seven years, right, to make sure that in particular SOC two audits are passed and any other regulatory and compliant needs are met, and then as far as the, other component, the ability to ingest public data but keep your private data behind a firewall and securely there is really what the specialization is that, so that’s the two part security conversation that an advisor could then tee up.

I’m gonna pause, we have a couple of questions that have come in, and we’ll take those from the chat really quick.

Let’s see here, how does it work commercially? What, what do you actually, what’s the product here, Brad? What is it that you’re paying for? Is it the platform? How is pricing set up? Is it I know you guys took a little bit of a unique approach to this, when it was launched. So if you wouldn’t mind getting into that for everybody so they can start to digest and wrap their heads around it.

Sure. So there’s a a platform fee. So we just talked about observability and all that, but the notion is you have one place to connect to that bridges you into a bunch of AI models, a bunch of data sources. So there’s platform fee, you know, dollars two thousand five hundred, dollars three thousand five hundred a month.

That includes a certain amount of AI usage. So Tom used the term tokens, but that’s just the metric of AI usage.

There’s a certain amount included with that, in the platform fee. Above that is a usage based system. So, you know, you could just basically have buckets of of AI brainpower usage.

That’s the platform in the AI side. On the data side, it’s about how much data do you want to host. So if you say, hey, the HR folks are gonna upload those hundred documents. That’s twenty gigs of data.

Okay. There’s there’s a per gigabyte fee in kind of hundred gig increments. And so the notion is let’s let’s get you some very simple things here. Let’s get some data in the system.

But as you decide, Hey, I want to add more and I want to add more and I want to add more, you’re adding recurring contracted revenue streams for data usage.

As you use more data, I guarantee you’re gonna be using more AI to process it, and it becomes this ladder effect so that you can turn a thirty five hundred dollars a month platform into a twenty thousand dollars a month MRR, you know, kind of total spend in a couple of years because as people start to inject more data and get more value out of it. So that’s the kind of laddering concept that we have going on that, you know, seems to be have good buying in the market.

And so what what I’m kinda hearing out of this is there’s it’s a fairly low entry point in comparison to what I think a lot of people think AI is gonna cost to get into it and build it out and do all the all the things, but it does have scalability for an adviser to see billing go up. And I believe you guys are, offer, like, account based on this, so, like, as the bill goes up, the advisor then rides along with that and gets Absolutely. Which is fantastic.

The the big piece that I’m hearing, and to answer the question that came up after this one is gonna be, you know, what industries, what market size, you know, in particular, is there any anything to attack? I kinda think this is across every vertical, right, on both sides. I mean, you can see SMB customers utilizing this all the way up through the enterprise, and I think you guys have already had success across multiple different, you know, manufacturing, legal, health care. But anything you guys could speak to in particular on those on that particular question?

I’ll do a quick one there. So, yeah, it’s definitely across industries, so there’s no real limit there. But kind of while the entry point’s low, sometimes this platform concept gets compared to, like, a seat license. I think somebody brought that up on chat. So if you’ve got, like, twenty employees, thirty employees, and you’re thinking, okay. Well, I could pay twenty dollars a seat to get this over here.

That might be tough for you to absorb the entry fee of the platform. So what we usually find is companies with a hundred or more people, it tends to be kind of a sizing thing so they can get past the the kind of that attempt to, make the equivalent kind of seat versus platform fee. And then that seems to be usually just like an early part of it.

And then folks that have large stores of proprietary data. So again, that’s not industry specific, but those are folks that there was an earlier question about how quickly do you get ROI. Well, that really is a reflexive of how quickly can you get the owners of that data, like Tom brought up HR, to copy paste from existing file store into the AI system.

As soon as you disintermediate, and a worker at the company getting access to the data from, you know, subject matter experts, there’s there’s workflows that you’re now changing and so efficiencies that you’re gaining. So the ROI, you know, within the first month, you’re you’re looking to pick off some workflows that have been very data oriented but have taken a lot of human time. And so that’s kind of where you kind of focus. I don’t know, Tom, if you had any thoughts on the industries or any of that.

No. Yeah. Everything you said, Brad, was correct. We are, you know, we are seeing a high, higher penetration, and the things we’re expedient is typically big in retail manufacturing. But across the board, this AI solution has opened doors for us to organizations, kind of in in every sector.

And it’s I I’m with Brad. We are ten we’re finding purchase in the companies that are, you know, over a hundred users. But as as you’re approaching a customer, if you have twenty users and you’re looking at, you know, OpenAI, you know, so ChatGPT Pro enterprise, that’s two hundred dollars a seat. Right?

So at that point, at twenty users, we’re still more effectively priced than that. So you really do need to kinda look to see what on the smaller SMB side, what is their what’s their plan to consume for AI? What’s their current, bucket? Because there’s still value even at smaller companies, but, you know, we’re finding a lot of success at a hundred users plus.

So there’s a couple other questions that came in, and thank you guys for those. I did miss the, the ROI one. Thanks for for attacking that one.

This one’s gonna be the easiest one, right? Hey, who do we contact at Expediant to kickstart?

And if we need to, we’ll follow-up with that, guys, and make sure that you have the right contact. It’ll be, Tom and Brad’s info, but along with the local, partner managers or a national manager that can step in and help if need be to route your opportunity to the right place.

I’m gonna skip down. Harry just threw one in, and I wanna attack this because I think it ties back in and we’ll answer the other ones.

So also, today’s AI seems to be driven around CX. Totally agree. I think five, six years of CX AI has been out there predictive and conversational. That’s been a driving factor, especially through advisors.

But he goes, and goes much further, what are the real tools and how do they help today?

We do know, we know they can, make current staff more efficient, what examples can lead to generate interest in clients? Particularly, I would I’m assuming here you meant from this side of the AI conversation, and, you know, Brad, you alluded to this a little bit earlier. This can actually be, a company, the CX conversation, right? You’re creating so let’s work through a scenario.

You’re in a deal selling the CX platform, and the AI that could be voice recognition or, you know, accent changes and and things like that that we’ve seen come out that are really, really cool on the on the CX side of things. But as you’re talking throughout the client, you’re like, hey. What are you guys doing for the rest of your AI conversation, or are you exploring other pieces of it? We can bring in a simplistic approach to take all of those call recordings, especially, like, say, it’s a health organization that has to hold on to those call recordings for a long time.

Right? Five, six, seven years, whatever that regular I think seven years. And so they have all these call recordings. You can now take that CX platform, dump it into this, have them analyze, and then give them back different, different recognitions of things that they could be missing from what could create a better customer experience, pull out commonalities, do better training internally.

On top of the stuff that Tom and Brad mentioned around HR, you also have legal comparison, right, where now people are utilizing legal search a little differently when it comes to AI. Here’s all of our common red lines in our contracts, here are some things that we can come into and start to streamline the use of our teams. Now, a lot of regulatory compliance around the legal side, but again, putting that in a secure platform that allows them to do what is understood to be within the guidelines and keep out what’s not, so it’s protecting the client from themselves, which is a number, one thing that a lot of organizations are, again, hesitant to utilize.

But Tom, Brad, do you guys have any other examples?

Those are the ones that just kinda flew off the top of my head of some of the use cases again that could be exciting for customers to to start utilizing the platform.

I mean, our own internal use case probably resonates heavily with people. So as we release the chat client, the very first thing you think crawl, walk, run, our crawl was our OSC reps. Right? We have ninety OSC reps.

They man the data centers. They man the tickets. They make sure that things get done. Their first use is that they just started using the chat manually for was exporting the their, ticket notes as a PDF and summarizing them, right, so that they can read these hundred page tickets easily and consume the data.

When we saw that, our run approach was we’ll build, I mean, our walk approach was we’ll build we’ll take a no code tool. We’ll use one of our connectors and connect it to the database, allow people to pull the ticket automatically by ticket number, and then get an executive summary and a technical summary. That found purchase in in, you know, sixty, seventy percent of our OSC staff. When you look at the hours that that saves that shift change, when you look at the hours that that saves that escalation, you know, the estimate is twenty five, twenty six hundred hours a year that we’re saving from this tool, which is, again, is something we’ve maybe invested fifteen hours into building the no code tool, getting it out there, and telling people how to use it.

That’s a hundred and sixty seven to a hundred and seventy thousand dollars of ROI that we see before we’ve really even productized the tool internally. So there’s lots of use cases around support where I think we’ll find purchase of tickets.

Let me summarize tickets. Let me get insight from all of the tickets we’ve taken. Let me ingest some of our playbooks and some of the understanding of how we solve problems to make lower and lower level tech supports more and more effective, freeing up the higher level techs to be able to do resolution cases and thus driving NPS scores up. Right? There’s an entire ecosystem here which one small solution being provided to drives efficiency across the organization.

So I’ll I’ll back up a little bit too. So is is and I think some of the comments in the chat have alluded to this. But the notion of a platform is valuable over time as you increase the kind of complexity in the AI world and the complexity in the data world. So, like, we talked about CX.

Well, hey, a CX vendor is gonna be really great within their silo in their world. But what happens when you need to correlate that to Salesforce? How does that happen? How do you desilo that?

Well, if you have a platform, you can now be connected to those multiple sources. So you can start asking questions across them. So that’s a a general sense of why platform is valuable, in concert. You know, not necessarily like, oh, I’ve gotta choose CX dot ai versus platform AI.

Just two different ways that are complimentary to think about it. But, specifically, like, a couple other, like, use cases. One, salespeople answer RFPs.

Marketing and sales enablement has built tons of information over the course of history of how to help a salesperson answer that. But we’re hiring salespeople all the time. Every company is. And so how do you ensure that all of that great information is indexed?

Oh, well, hey. The marketing team has a CMS, and then you gotta figure out where to log into, and then you gotta ask the question the right way or which year was the all that’s just intermediated. Now you can go into AI chat and say, hey, I have an RFP. Here’s the twenty questions.

What’s the answer?

Here’s the answer for question one. Here’s the answer for question two, and here’s the link to the source material. So essentially, you’ve you’ve made it easier for a salesperson to respond to RFPs, get the right marketing material. That’s a pretty big one. Another one that we find people doing is, just onboarding and enablement.

So constantly hiring new employees, they go through different tracks. You have an LMS that you put them through. But all the information that’s behind the LMS, it’s like, Hey, you go through it in your first thirty days, but what happens when you have questions later? And so, you know, you can, again, dump that into a file folder, and now it’s AI accessible. So you can sit there and be like, enablement and onboarding doesn’t need to stop in your first thirty days because your questions don’t stop in your first thirty days. So people are thinking like, hey, what’s all of this enterprise data we’ve created that’s unique to our organization that we can now make AI accessible to employees and applications?

So those are a couple additional ones that we see people kind of popping to, but definitely things that can improve sales and the ability to answer sales questions, you know, very there’s a lot of motivation around making sure that that that’s accessible.

I had something just kinda dawn on me, and this is, this is always dangerous, right, just having an original thought in front of a live audience, but, it sounds like what you guys have built with this platform is similar to a data fabric that did interconnection into, you know, multiple different things for, like, cloud provider, like AWS going into, you know, different, like, a RingCentral platform and connecting privately and then securing it a little differently and allowing data to transport on a data fabric. This is almost like an AI fabric. You guys are connecting into multiple different AI platforms, allowing the goodness of the technology to flow to the customer, but in a secure manner in which none of their data then reaches back out publicly.

So, there’s a couple of really good questions, but I just, as I wanted to describe this a little differently, like if I was sitting in front of a client, I’d be like, these guys have built an AI platform that’s like a fabric into all the different LLMs, and somebody had asked, and we mentioned it a little bit earlier, Tom, but can you hit on the major platforms that this can connect into? And I think it’s endless, right? You guys can connect into any of them, including, and it’s not a silly question, the Five9 and Genesys platforms of their AI, if they have a platform that’s allowing connection, you guys can go into that and maybe tackle both of those two questions out of the chat if you don’t mind.

Yep, so initially, ChatGPT, so all of their models from OpenAI, all of the models from Anthropic, which are the Claude models, all of the models from Google, all of the, llama models, and then most of the models that perplexity solves, provides. The the the limit of the models isn’t based on models we that today we have. They are vendors who are willing to attest publicly that they will not train on your data, and they’re SOC two compliant and HIPAA compliant. So there are vendors who can bring up every single one of the eighteen hundred open source models, and we can include those into our system when clients need them. They have to have an understanding of how the model was trained and what the biases are in that model. But as long as the vendor is not training on the data, not, sharing personal data, not storing it, and they’re HIPAA compliant and and piece and, SOC two compliant, we can bring them into our platform.

So let me let me kinda walk through a a scenario that I think could happen or could help with the advisers. So you got a customer. First question, this is gonna go into what are some questions and talk tracks. I’m gonna throw a couple out, and I’d love to get your guys’ opinion, but essentially, the number one question is what are you doing, what does your AI roadmap look like, or what kind of pressure are you getting to implement AI and what are you doing with it?

If you’ve got companies that wanna go to Google, all of these guys now have vertical centric, you know, LLMs that you can go, like Google has a healthcare LLM out of the box. There’s a financial LLM out of the box now that you can, because companies can go directly to it, but what we’re saying is, hey, instead of doing that, route through Expedience platform, do it, and here’s how they’re gonna make life easier for you, and monitor this, and especially if you have a resource starved team or teams that you’re just not quite up to date on that protect your own organization, this is a really nice security measure and blanket to offer at a relatively low rate, because you’re going to be getting into the same information, but now you’re doing so a lot more responsibly from an organization’s viewpoint, so that’s a really good key talk track in how to introduce the idea, in my opinion.

You know, what are some other key questions that you would ask, you know, Brad or Tom to to kick start the conversation, or what have you seen? And then we’ve got about three minutes left, so I wanna save at least one minute for that quick win, Tom, and and then you can take us through.

Sure. So, yeah, some of the easy questions to ask to get these kicked off are, you know, have you implemented, any part of your AI strategy yet? It’s a simple question, but where are you in that process as opposed to even the road map? Right?

At some point, there’s some AI, choices already made in every organization. Where are you in your AI implementation strategy? What are the long term goals of your AI implementation strategy? How are you handling governance?

Governance, GRC is going to be the talk of twenty twenty five. How do you handle risk, compliance, and governance around AI? That question hasn’t been answered by the big providers yet, which is the place where we are gonna find we as, TAs, we as Expedient are gonna find the most value that we can drive to customers. It it’s amazing to connect the data, and that’s where they’re going to see ROI is in the connecting the data kind of ecosphere of bringing in data.

And there was a question about copy and paste. It’s not copy and paste. It’s just bringing that data down into a new repository, having that data live. There is some data duplication issues, but we can talk about those separately.

But, really, it is how are you handling governance around AI? What are you doing for enterprise search today? What are you doing for AI? Those are the questions that are leading to further discussions.

And and just to tackle this really quickly, and it was gonna be a short answer, we can expand on it in a bit, but this is different from Five9 and Genesys. Genesys and Five9 are gonna be very specific AI that’s gonna drive the customer experience. This is a platform that can extend into that, and help support it and grow, but this is gonna be on that other side of the generative, I can never say it, GenAI, it’s a running joke around here, I can never say it right, but this is what gives you the ability to utilize the ChatGPT and mix in your private data with that, and some of that data could be, originating from your platforms of Five9, Genesys, etcetera, through your customer experience, that can also be pulled in, so this is more on top of, and in addition to, versus like a competitive balance there.

The other question here that I wanted to ask, let’s see here, Tom, and then, we’re gonna go over one minute, Doug, sorry, but I wanna hear that story, but I wanna get this question answered. How do you keep the customer’s production data synchronized with expedience data?

Can you give a quick, you know, overview of that?

Sure. That is just a standard ETL process. So you’re not moving the data over. You’re vectorizing the data.

So you’re pulling it. You’re pulling out the text. You’re vectorizing. You’re embedding the semantic search, values into that data, and then that data is kept in a new, a new environment for that customer.

And the customer actually owns that data. We don’t own that data. So what you’re creating is a new source of data, vectorized data for each of their data repositories that that the customer can utilize themselves. Five years from now, that that all data will most likely be vectorized, and this just starts them on that path.

But there there’ll be a a synchronization frequency. So as data gets added to wherever it originates from, you know, you can kind of look nightly and say what’s the changes, and then we, you know, kind of update the the the repository based on what’s changed.

It’s almost like a backup copy. Right? You’re just you’re just updating the changes, not the you’re not repeating the whole process over and over again?

Yep. You got it.

Alright. So so tell me the fastest that you can deliver it. Can you take us through that really good win story? I think it was a a couple of calls and a close. It it sounds like something that would be really impactful.

Yeah. So Luna came to us through one of, our TAs, tier four. And, tier four, brought them with the understanding of they want to explore AI, but they weren’t sure what they wanted to do. Very traditional to almost how all of these calls start.

And in the discussions, it very quickly went to, if I can provide compliant chat and private chat to our internal employees right now and give them access to it, lower the risk of shadow AI, lower the risk that they’re leaking our data or leaking our performance, anything that they’re pushing through AI today. I can give them those productivity tools. The true, like, kind of crawl win, the it immediately led into, hold on. Explain to me how in one day I can connect my SharePoint data sources in and automatically solve this enterprise search problem that we’ve had across you know, one of them is HR data directly.

One of them for them is the kind of RFP use case we use. And in both of those cases, everything that they had thinking about long term AI became ancillary. Like, awesome. You’re gonna give us a platform.

We can think about that next year. For right now, let’s get AI in, get all of our users more productive, and then let’s start getting ROI out of enterprise search. This exact story we tell took ten days to close from introduction to them signing the paperwork. The TA hadn’t seen that kind of speed.

This is just the world we kinda live in today with these closest.

It’s releasing that pressure. Right? Yeah. It’s mounting on these organizations. Guys, thank you very much. I think this was super insightful and gives us a new view on how to tackle an LLM conversation for advisors, and there’s already been a lot of success with it, so thanks again for joining, Doug. We will hand it back over to you, sir.