Shawn Cordner (00:13.454)
Hey everyone, welcome back to the Amplitude of Tech Podcast. I’m Shawn Cordner, Chief Marketing Officer of Amplix. Today I’ve got a great episode with Robby Gulri. He’s the channel CTO at RapidScale. Robby’s a cool guy. He’s really smart. He’s got a lot of interesting things to say. His job puts him in front of technology leaders from all over the world, all sorts of different businesses. And so he’s got firsthand knowledge of how people are leveraging AI in their businesses and helping them leverage AI to its fullest extent.

Super interesting conversation. Hope you enjoy it as much as I do.

Shawn Cordner (00:51.912)
Thanks for joining me today. are you? It’s been a long time. I was trying to think back to how we met and I believe it was because my former branding agency that I had, we did the rapid scale rebrand and after we did that project, we had some new messaging and a new logo that we wanted to introduce and so I came to your SKO. Is that right?

How are you? It’s been a while.

Robby Gulri (01:16.054)
That’s exactly right. I was really happy to see you there. And we were talking a lot about messaging, how important messaging is, how important, frankly, storytelling is. And so I think that’s how we sort of hit it off. And here we are trying to do a podcast together. So I’m looking forward to it.

Yeah, I think that’s how we bonded, right? It’s not often that I run into people in the tech space that are not in marketing, but that speak the marketing language. You had a little history in marketing, I believe. Maybe just give us a little bit about your background and then what you’re doing today for RapidScale. Yeah.

Thanks, Shawn. A little background, man. I’m to go all the way back when I was studying in college. I went to Georgia Tech, studied engineering and mathematics and thought I was going to be a hardcore engineer. And I tried that for a little while and I absolutely hated it. What I had found out that I love to do is to really think about how businesses solve problems with technology. So I became a solutions engineer, which led me to, you know, sort of product marketing for a little while and inevitably…

I’m one of the few people that I’ve run into that has held a CMO title, chief marketing officer title and a head of technology title. So I like to pride myself on understanding the technology, but more importantly, basically trying to be able to articulate the value of technology to generate some kind of business outcomes. And that’s fundamentally what I do at RapidScale. I’m the channel CTO and you know, fancy title and fancy words to ultimately do what I love to do, which is to go out, talk to CIOs, CTOs.

practitioners of IT understand their biggest challenges and try to figure out how we and us along with our ecosystem of partners can ultimately solve those top challenges that they’re willing to.

Shawn Cordner (02:52.258)
Yeah. As a CMO of Amplix, I can tell you they hand out that title to anybody, but CTO, that’s title that you earned. That’s actually one of the main reasons that I wanted to talk to you. We hit it off and I think we bonded, like I said, over the marketing stuff, but since then you’ve come out to a number of our events and you participated in some of our webinars. We’ve stayed in touch. Your position as CTO there, puts you in front of a lot of different CIOs.

that are exploring a lot of different use cases for technology and specifically around AI. so I thought, it’d be really interesting guests to have, because you could talk about some of the coolest use cases that are out there for AI as, know, people are trying to navigate their, their journey, into AI and, and, I think it would be interesting for people to hear what their peers are doing. But before we get into that, I just want to talk like,

We had a little prep call about this and you were talking about the promise of value in AI. So let’s talk about that first. What do mean by the promise of value in AI?

There’s lot of opportunity at the same time. There’s also a lot of general misconceptions. And we’ll elaborate on all of that. I think there’s definitely promise. I think there’s definitely opportunity. I think there’s definitely, you know, the particular opportunities to leverage data in ways that we’ve never really been able to use data to generate insights for the business. So that’s the promise. Unfortunately, what I’m seeing and what we’re seeing out in the industry is a lot of organizations.

have the right intentions, they have the right ideas and they have the right kind of potentially some internal alliances in play. They don’t have as often a data strategy to get them to what this promise land essentially don’t like.

Shawn Cordner (04:45.17)
I think people, it’s a big apple to start taking bites out of, right? So there’s some fundamental foundational things that need to be addressed. And I think we’ll kind of circle back and talk to that a little bit. at its core, a lot of people consider build versus buy. How are we going to get started? How are we going to get, know, dip our toe into the AI space? And a lot of times it’s buy but…

I think you see a lot of the build use cases, right? So maybe talk to me a little bit about the build versus buy and what’s the value of starting with one versus the other.

Yeah, yeah. And before I get into the use cases, let me kind of, you know, again, I’m in the luxurious position of going around and talking to really smart people, trying to do really interesting things for their business. And ultimately there’s three fundamental categories where a lot of these conversations are kind of taking shape. Number one is this notion of hybrid cloud, right? Is, is can I, you know, make sure that the right workloads fit in the right cloud? It doesn’t necessarily have to be just public cloud, or it could be…

you know, a number of variations around that. So the notion of hybrid cloud optimization is something that CIOs are really thinking about. Cyber resilience is another area that they’re definitely doubling down on. Can I be reactive, more proactive and reactive with my cyber postures? And then thirdly, which we’re obviously going to focus in on is sort of data readiness and AI. And before we get into AI, it’s really around data readiness.

What, what I’m seeing is organizations finally starting to think through, okay, what the right data strategy needs to look like in order for me to activate my AI projects. And ultimately there’s a lot of use cases that we can get into at RapidScale. We’ve built hundreds of use cases for different industries, different sizes. But ultimately it often these conversations that inevitably start with AI first have to.

Robby Gulri (06:34.604)
kind of basically align themselves from the data readiness, data strategy, data lakes, single source of truth, a lot of things that we can cover around the data. But that’s honestly the first magic sort of ingredient when it comes to a lot of these AI projects.

Yeah, I think when you’re talking about data readiness, one of the key things that’s a concern is data quality and trust. Can you trust the data that you have? The AI model is only going to be as good as the data that you put into it. How is it that a technology leader can know whether or not their data quality is good and ready for an AI model?

I spend so much time, Shawn, working with clients, really defining, helping them define what data is. Often people think data is in a database, which is true. That’s absolutely true. But there’s data in other places. There’s structured data, there’s unstructured data, there’s transcripts of YouTube videos that are potentially leverageable. There’s PDF files. We did a project actually in the Northeast with a lobster company where the data was frankly files, literally physical files.

sitting in, you know, those vanilla folders and that was their version of data. So how do you gain, you know, insights and digitize that data to be able to generate some models, right? So the first is definition of data. So that’s, that’s critical is to make sure that the organization understands that, that, that what data is available to them. Number two is, is, is

Often organizations fail to realize that there’s third party data that’s available that can be leveraged for whatever use cases that they’re trying to, trying to basically solve for. I’ll give you an example. We work with a logistics company. Logistics companies got tremendous amounts of data. I mean, petabytes of data from IoT sensors that they’re basically capturing. Ultimately, what they’re trying to solve for is a prediction use case. And I predict when I need to send the Army Corps of Engineers to move sediment around.

Robby Gulri (08:27.896)
So can basically guide the ship through the Mississippi River at a surface level. That’s a pretty easy thing to understand. Instead of doing it mechanically or manually, mean, leverage the AI that I have to be able to, or the data that I have, be able to then, you know, make that decision and then finally send the only floor of engineers. However, in that use case, we had over 50 data sets that we were leveraging, right? Some of them were internal data sources. Others are external.

You know, the currents, the traffic, you know, within the Mississippi river, all of those external data points and data sets basically, you know, impact the math that basically is created for, you know, for solving this particular use case. So it’s really a couple of things. Number one is the, you know, single source of truth. How do I get data across a multitude of sources into a common data lake? That’s number one.

Number two is what third party sources of data. And by the way, when you start talking about third party sources of data, you’re thinking about not only CSV files or scrapes of websites, but also APIs into third party sources. Some of these APIs are free. Some of them are available on GitHub. Some of them are, you know, potentially procureable, you know, through third party sources. There’s brokers of data all over the place that bring a lot of interesting data elements to you. So that’s one of the things that we try to bring to the table is basically a holistic understanding of kind of the data.

universe to be able to ultimately help these organizations solve these challenging problems.

Yeah, I think everything is data these days, right? Right down to conversations that you’re having on zoom and there’s AI transcripts and your contact interactions, whether it be through chat or whether it be through voice conversations, it’s all data now. How do you kind of sort through and prioritize the data that’s available to you? How do you know what data is relevant? mean,

Shawn Cordner (10:17.494)
You must have to start with an end goal in mind or a specific project in mind and kind of work backwards. What are the relevant, you know, bits of data that are available to me within my organization? And then where can I go to third parties to enrich that data?

Yeah. I think the first question you have to start and if you’re kind of, you’re a listener and you’ve run an enterprise or you run an IT organization, you run an enterprise, ask yourself a simple question. Do you have a data culture? What I mean by that is we talk a lot about general culture, but the data culture is really important. It’s frankly, frankly, the most critical, right? Organizations, even in 2025, don’t know necessarily what the value and how valuable that asset is. You know, we’ve heard data is the new oil.

But a lot of organizations may say that, but they don’t necessarily have, you know, an encouraging culture around that continuous sort of iteration of data. Right. So number one is let’s figure out how to basically align the organization from a data culture perspective. Right. And then it’s, it’s, you know, you talked about, um, sort of the governance side is, is where is that data? What, know, does it sit, does it sit internally with me? Is it, is it on a SharePoint? Is it on OneDrive? You know, what are the different data sources? So let’s actually define where those data sources are.

What that does is it gives me a little bit of an analysis and a utilization of those data sources. So that’s kind of the next step, right? Is let’s go from data culture to data analysis and utilization to now start thinking about integration and collection. Ultimately, what our goal is, is to, from a data perspective, is to get it in one common platform, one data lake, one sort of store of some kind.

Back to what you said, I love what you said. You said all this is unnecessary and frankly useless unless you have a business outcome in mind. That’s ultimately where the first place to start is what it is, what is it that you’re trying to accomplish? What is it that, you know, ultimately is going to move the needle for your business? What’s going to save you money? What’s going to generate revenue? What’s going to basically have a defined ROI? Once you understand that particular business outcome that you have in mind, all the data alignment culture stuff that I talked about is going to be pretty.

Robby Gulri (12:22.709)
it’s going to fall in place.

Yeah, but conversely, think that a business should treat all of their data as if it’s going to be used in the future, right? Because what you don’t want to have to do is delay a project in the future because you have to go back and clean up a bunch of data when you can draw a line in the sand today and say, we’re going to start to assume that all of this data is going to be used. Is that what you mean by data culture?

Yeah, absolutely. Make the assumption that all your data is going to be used at some point. Maybe it may not be, but also what I mean by data culture is having, it’s honestly organizational alignment around the data. Meaning if you’re a data analyst, of course your job is to capture, secure, govern, and make sure that you generate insights from the data. But man, there’s a lot of people outside of just the data analytics group and the organization. So making sure marketing understands that there’s a single repository of the data.

making sure that sales understands that there are Salesforce, but there’s other data that we may be able to be able to leverage, right? So having the leaders within those business units really align themselves around making sure that A data is a valuable asset as a defined asset, and then being able to, you know, again, leverage it for the things that we want to accomplish.

Yeah. One thing that comes to mind with data culture, you know, being in marketing, we have to pull a lot of reports to show that we’ve got a return on the investments that we’re making. And the constant struggle for us is aligning our activities with actual sales numbers. Usually where the gap is, is with salespeople. We have great salespeople, they’re professionals, but there’s not always great discipline. This historically speaking, I’m not just talking about amplates, just in my career, there’s not always great.

Shawn Cordner (14:03.884)
discipline with salespeople in how they’re entering data into a CRM or if they’re forecasting accurately and, and that kind of impacts the output of the data that we’re able to create. so just, you know, when you, when you were talking about kind of aligning culture and aligning the people, users, and how they use data to me, part of that is creating an understanding within the organization that this company is.

moving towards AI and we’re going to be using this data as inputs to create outcomes that are going to benefit all of us and create more value for the enterprise. so to do that, we need you to behave in a certain way. And that’s the reason why I guess I, what I’m getting to is communicating the why and not just the what.

Yeah, I think there’s some fundamental rule of thumb is that you can kind of apply, you know, what you’re talking about is what I’ll call transactional data, right? And the word transaction was not intended to be, you know, disparaging anyway. This is important. This is the lifeblood of the business, right? This is sales transactions, inventory records, customer service calls. Like that’s the lifeblood of the company. That’s the transactional side of it. Day to day, you know, how often does this data change? Where am I capturing it? All the, you know, the four V’s of data, if you will, right?

The next layer is kind of behavioral data. We’re trying to do, I don’t know, some kind of sentiment analysis, right? For in marketing, we do a lot of that. Behavioral data. We’re talking about web clicks. We’re talking about downloads. We’re talking about, Hey, how is this particular prospect or customer engaging with me and whatever platform that basically I’m trying to engage with? That’s really what behavioral data is. And then, you know, like I said, we talked about a little bit of the notion of third party data.

external resources or sources of data, research reports, demographic information, competitive intelligence, and, know, various other sort of, broader sort of market perspective data can be also brought into, you know, some of the use cases that we’re ultimately trying to solve. So you’re absolutely right. It’s the culture around data, but then the definition of data to be able to then categorize the data and be able to, again, solve for a business outcome. The AI really comes after that’s the after effects, right?

Robby Gulri (16:15.05)
Once you get the foundational stuff in play from a data perspective, that’s where, you know, ultimately AI can make sense for you.

When you talk about the definition of data, this is a question just out of my ignorance because I don’t operate at the level that you operate at. Does IT necessarily know all of the data that’s being generated and collected within the enterprise?

generally, I’m generalizing, generally the answer is no. Like, like we, us tech guys like to say that, like we have a view or visibility. think, I think we’ve made a lot of leaps and bounds and progress over the last 30 years of me basically being in technology. I remember God years and years ago when I was early in my career, you know, we always talked about, man, the CIO has got to have, you know, a seat at the table because the CFO does, the COO does.

But the CIO has always looked upon as the fixer or go, you know, run my systems or change my password or whatever. Never really thought about as a strategic, you know, aligned organization. I think a lot of that has changed. That culture in most organizations has changed. They realize that technology is not a department. It’s actually immersed and embedded across every department inside the organization. So the CIOs and the CTOs sort of statured.

in the business has obviously been elevated over the last 30 years. I think that’s a good thing. I think also though, there’s the assumption that the CIO has a viewpoint of everything that’s been in the business. Maybe ideally they should, but we know this. Shawn, you and I have spent some time in marketing. How much does marketing leaders, how much money or investment do marketing leaders spend on IT technology?

Robby Gulri (18:04.492)
Whether it’s HubSpot or CRM platforms or I don’t know, social media engagement tools. There’s a lot of technology behind the scenes to basically do marketing well. So, you know, the IT person, the leader may not know all of that. And guess what? All of those sources of data or all of those sources of applications generate some level of data. It could be transactional, it could be, you know, more behavioral, but regardless, that’s a data element that IT people generally don’t have, you know, completely disavowed.

Yeah, I think that’s what I was thinking is complete visibility, right? Because IT might be aware that marketing has HubSpot or Sprinkler, right? But they might not know what data is being generated and collected within that application. And that could be data that could be used to enrich other data sets, right?

That’s exactly right. You know, at the end of the day, what we’re trying to get to is an answer of some kind, right? Should I do something that I’m not based on the analytics and the machine learning and some of the AI that I basically could apply to the data that I own combined with the data that I may not own, but I have to, right? Can I, can I personalize a service for you Shawn, because versus somebody else, because I know something about you because I’ve collected

the right amount of data. You know, we’ve got a lot of intent tools that we use actually internally, and we actually allow the partner community of ours to use some of those intent tools. Zoom info demand base. These are awesome platforms that try to basically forecast and predict, you know, certain engagement metrics and certain.

propensity to talk to people based on the data that they’ve captured. so a lot of organizations internally are trying to, you know, basically create these personalization engines based on the data that they have. The third area that I would talk about when I talk about sort of the applicability of data in the context of business is what I’ll refer to as segmentation. I cluster? Can I bucketize, if that’s a word, my customers in groups? So basically…

Robby Gulri (20:02.39)
allowing, you know, group A to act differently or me to target groups A differently than maybe group two or group D and group C. Can I basically create market and micro segmentation based on the data? So can I identify neat segments that I may not necessarily know without really analyzing that? So there’s a, there’s a lot of things behind the scenes that can be done. Assuming again, we’re talking a lot about data today, but assuming that the data strategy and the data alignment and the single source is true.

is in play, is in place. Once it’s in place, then we can do a lot of interesting things for the business.

Absolutely. You know that when you, when you talk about segmentation and you talk about personalization, I think that that kind of triangulates with data enrichment as well, right? Those are kind of three legs to a stool. Because you can create micro segments and that leads to the ability to create personalization based off of attributes of those segments. But oftentimes you’re going to want to enrich the data.

to, you know, with maybe third party data in order to kind of feed that model. Right. So can you talk about kind of those, those three points and how they work together? What’s the interplay?

Yeah, yeah, that’s a great question. wow. Let’s talk about data enrichment first, right? So let’s just define what data enrichment is at a high level. data enrichment really is, filling out the gaps that may be incomplete in the data that we’ve captured. Right. So can I, can I, let’s use a very rudimentary example. I’ve, I bought a list from Sertz Hardy Booker of some kind. within the list, I’ve got the first name, last name. I’ve got the mobile phone number. I’ve got the email address.

Robby Gulri (21:45.442)
But I may not have the domain of the company that he or she works for. Right. So what are we doing? We’re basically filling the gaps. We’re filling that domain name because that may help us in the future to basically do some kind of intent data mapping or whatever. So there’s a main name filling in that particular field in the, in the table is a form of data enrichment. Right. There may be other forms of data enrichment. Let’s say I captured an image, but the, you know, a little bit of pixelization on the image.

Can I apply, you know, basically some data enrichment techniques to get that image in a place where I can actually generate some insights for it. So that’s what I’m talking about when we talk about sort of conceptually what data enrichment is. How it sort of applies to personalization and segmentation data enrichment. You know, what we’ve talked about the data platform and the data lakes and the single source of truth. Now you’ve got a layer of data that you’ve enriched and you’ve created a much more complete set of data. Now we start applying some of the

interesting business cases and news cases that can be applied to this considered data. So the reason it’s a triangle, the reason it’s all sort of part of this convergence model is without some accurate and rich data, I cannot create a personalization type of use case. So give you an example. let’s say I’m trying to do, I don’t know, we all go on Netflix and, you know, basically occasionally we’ll take what Netflix has recommended to us and we’ll go watch something that, you know, maybe weren’t, wasn’t on our radar.

because the personalization engine gave us some recommendations. That’s an example of data that basically was enriched behind the scenes to say, Shawn, you you like documentaries based on what you like in your behavior and some of the things that you’ve sort of, we’ve learned from you or about you, I’m going to recommend these kind of three things. So that’s an example of personalization, but it has to start with, you know, the data enrichment side. Let’s say I’m trying to do adaptive pricing. Let’s use more of a real business example, right? And I want to basically create personalized incentives.

So I’ve got loyalty program in my retail organization. I want to create, you know, basically a personalized incentive plan or, or, or some kind of credit limit, you know, based on what I know about you and your spending behaviors. Guess what? In order for me to get to the right risk level and the, and the, and the right type of, you know, sort of statistical modeling that I can live with, I need to be able to have an enriched data. So that’s where kind of personalization, you know, sort of feeds into play, right? Segmentation is the same thing. You know, if I’m trying to do risk tiers.

Robby Gulri (24:06.242)
And I’m trying to basically segment my merchants. don’t know, maybe I’m doing claims and fraud analysis, you know, an abuse risk sort of mapping. If I don’t have completed and rich data, then I can’t really come to a differentiated, you know, sort of, you know, approach here and also, right? So all three of those things, and I’ll actually add prediction falls in the same kind of realm, right? We talked about personalization and segmentation, predictions are the same. I can’t accurately predict within a certain level of standard deviation.

you know, anomaly detection or some kind of time to event or some kind of risk scoring if I don’t have accurate and enriched data. I can go on and on, but I’ll stop there.

No, that’s all really interesting use cases. But it kind of leads us to the dark side of this conversation, which is if you’re doing all this personalization and this segmentation and even prediction, and you’re basing a lot of it off of third-party data that comes from marketplaces, we need to talk about privacy and we need to talk about compliance, right? Regulatory.

100%. We are not, we are not, I’m putting myself on do not disturb us. I’m getting called on use two seconds. Yeah, we got to talk about governance. We got to talk about privacy. We got to talk about, you know, a lot of different use cases. Let’s start, let’s start with what a lot of CIOs are concerned about right now. You know, right now, 50 % of organizations in America, think a recent MIT survey showed this. Employees in those organizations are actively using chat GPT.

They’re actively using Gemini or any other perplexity, any other tools that, you know, we’re used to. So, well, inadvertently, they’re, they’re, um, releasing a lot of potential company IPs, you know, uh, they may not even know what they’re doing. They may ask a simple question that is pretty innocent or seemingly innocent, but it’s training the models on my proprietary data and information and insights about my business. Maybe it’s for my competitors to take advantage of, right? So a lot of PIOs are thinking through, okay, how do I protect that, that, that

Robby Gulri (26:09.4)
query that prompt that, you know, my employees are basically ingesting, you know, injecting into the, into the public engines of the LLMs. And so there’s a number of tools available. And in fact, we just did a deal with a company called Liminal and we’ll be releasing that here pretty shortly. But ultimately think of it as a proxy. Think of all the prompts to the LLMs going through a central place. It’s audited. You have to authenticate to it. It’s got to be integrated to an active directory.

So now it gives me as an IT operator, as a leader of an IT organization, to basically say, yep, I can basically trace where all of the queries are coming from, not only at the user level, but I think group level. I can also look at trends. If an employee base is saying, teach me how to build a resume, a bit of an employee sentiment analysis tool that I got there, right? So there’s a lot of things that are happening on that side of the house. So that’s an area of privacy that organizations are thinking about. The other area when it comes to data.

is especially in heavily regulated industries, healthcare, financial services, et cetera, is, you know, what am I doing to, you know, redact or anonymize the data in ways that it can still be leverageable, still be used to, you know, generate the outcomes that I want, but ultimately I’m not exposing any kind of PII or PHI on a healthcare organization. So there’s a lot of conversations, technology and governance sort of compliance.

frameworks that are being applied to those kinds of things as well. I don’t know there’s a perfect answer or a perfect solutions, but there’s definitely more and more thought and conversations and boardroom conversations around, you know, the topic of AI integrity, AI ethics is obviously as well as AI security. And what we haven’t talked about and probably don’t have time to talk about the same tools that we’re talking about that are hopefully adding some level of productivity and efficiencies to organizations are the same level of tools.

that the bad actors are using and frankly have access to as well. So what we’ve noticed on the security side is, you know, phishing attacks being much more targeted, segmented, personalized, all the things that we talked about from a positive business use case perspective. We’ve talked, we’ve seen, we’ve seen the volume of these attacks much more elevated over the last year or so, because now it doesn’t take a lot of human labor for me to…

Robby Gulri (28:33.058)
you know, launch these zero day sort of attacks, I can do it with a handful of agents, know, AI agents that are basically running behind the scenes.

Yeah. AI is definitely made at the wild wild west when it comes to security and the tax break. I just want to go back to, you I think what I’m thinking about is as an enterprise, you need to have kind of a culture of transparency and explainability when it comes to data and how you’re using data and privacy of the people whose data you’re collecting and leveraging.

And I, know, GDPR and I’m pretty sure CCPA require you to, to be able to tell someone if they requested how you’re using their data. so just going back to an example that you made Netflix and I don’t know this to be the case with Netflix. So please don’t sue me, but I can imagine a world where, you know, their recommendation engine is not just the data that they’ve collected on your user behavior, but also third party data about your user behavior across.

other platforms, right? So they can understand that you’re interested in what your buying habits are and things about your demographics and psychographics that they might not have access to specifically on their platform. so, you know, if you’re a company that’s enriching your data with third party data, you should tell people that and you should tell them how it’s being used. you’re a company that’s collecting data and that data to third parties, you should talk about that. You should tell people about that. mean, talk to me a little bit about that.

Yeah, I don’t know either. That’s a great point about Netflix. I’m assuming this, know, they’re like leveraging data in so many ways that we don’t even know. All we see is the personalization engine, right? I know, I read a white paper a while ago, they were talking about like, even when we pause, like that’s important, like evidence information. Even when we rewind to something that we’re watching over and over again.

Robby Gulri (30:33.826)
That’s important information. So I got to think that they’re doing some interesting things from a data perspective. But I think, think the key principles, let’s talk about again, foundationally, right? The key principle is the principles of data privacy. Whether you’re talking about GDPR or, or PCI or Sarbanes-Oxley or name the, you know, the hundreds of local state, national, global, you know, sort of compliance frameworks. Ultimately, number one is the fundamental principle is to dissent, right? So, so, you know,

Absolutely right. means that the organization must be explicit in the way that they’re informing their constituents, getting permission from individuals if we’re talking about consumer or your businesses, making sure that the data is collected and processed in a personal way. We are very proud, at RapidScale, we’re very proud that we are SOC 2, Type 1, Type 2, because it proves…

And those reports are available to anybody who asks. It proves that we are doing everything that we can and we’re not just checking a box. We’ve got the checks and balances to make sure that if you are running your platforms on our cloud infrastructure, they it as being, you know, we are doing what we can to manage and secure and govern that data accordingly. Right? So that’s part of our journey as an organization. So consent is definitely, you know, a fundamental principle. Then you get into sort of, let’s call it purpose limitation.

Right. What am I using this data for? I’m going back to the Netflix example. What is it? Is there a legitimate purpose? Is there a, is there a, am I just collecting it for the future? You know, and basically maybe saying that it’s going to be oil for me in the future. There’s got to be some level of light purpose limitation. If you look at any of the terms in service of companies, legitimate companies, they will always have a purpose limitation, you know, sort of claws in it to say, Hey, listen, we are collecting your data, but explicitly Shawn.

We’re going to use it for A, B and C purposes. And the third thing, Europe is bigger in this, frankly, a lot more mature when it comes to GDPR-oriented kind of policies that America is today. This sort of data minimization. What’s the minimum viable data, the MVD, as I’ll call it, data that we need to capture to be able to still do what we need to do as a business, but not just overly expose any of our constituents, right? So that’s an important element as well.

Robby Gulri (32:51.35)
which then leads into accuracy, quality, security, you know, the technology, actually this is less for me. What I’ve seen is less about technologies because there’s, there’s ways to encrypt, right? Encryption techniques, protocols, you know, I spent 10, 11 years in cryptography. know, I know encryption pretty well. Pretty hard, you know, protocols and standards around that. It’s less about the technology and more about sort of the capture, storage.

and consent and use of this data that I think we all have to ask those questions, hard questions to our organizations around to be able to get the answers that we want.

Yeah, absolutely. you know, here at Amplix, we’re not doing business with, anyone in Europe. we don’t expect that people from Europe are going to be visiting our website. but we do maintain GDPR compliance because our core value is that we’re going to be client centric. GDPR is, is the most, consumer centric, regulatory, know, statute out there when it comes to data use and privacy and things like that. So.

You know, we just thought it was consistent with our values to adopt that and just kind of take that position that we wanted to carry that standard forward. I know that’s probably expensive for larger enterprises, but you know, d is there something to say for, making that a strategic message point since we’re we’ve talked about the intersection of, marketing and technology a little bit here. you know, is, is that something that can add to the brand and give them a strategic position in the.

I think so. It depends on how it’s done. know, a lot of people throw these terms around without the why and a very rudimentary or fundamental understanding of like, or an explanation, not an understanding, an explanation of why this is important, you know, for whoever that I’m targeting. If I’m talking B2C, B2B or, or, you know, some other hybrid of that.

Robby Gulri (34:49.038)
So I feel like there’s definitely an opportunity to create value, create a brand, create an identity around privacy and security. I noticed it, you’re talking about Europe and GDPR. You’re right. It’s arguably the most stringent privacy law that I know of. I noticed it as a consumer. I spent a couple of weeks in Europe this summer. My wife and I went to Scotland and England. even when logging into my own Netflix account in Europe, I saw a pop-up that I never see here in America.

When I was at a hotel and I turned on the TV and you know, the info entertainment system had the whole like privacy statement that I don’t see when I stay at a Marriott, you know, in America. So absolutely there’s an opportunity to kind of use those touch points in ways that, you know, maybe organizations in North America aren’t necessarily doing today.

I’m going to pivot away from data because we just spent a lot of time talking about the least sexy part of technology. But I want to go back to something that you said earlier. You were talking about how, you know, a CIO should have a seat at the table in business discussions. And I think that reflects, you know, in my career, 27 years now, I’ve seen a change from IT being and, and contact centers really being a cost center to being a value center. And I think, you know, the CIO.

having a seat at the table in those business discussions is a reflection of that. And I know that when we had a prep call, you were talking about a book that you’re in the process of writing that I think talks about that subject. So let’s just spend a few minutes talking about the journey that technology has taken in the enterprise from being a call center to a value center.

These are hard lessons that I’ve learned. And thank you, Shawn, for giving me this platform to talk about just the journey in general, As like, I’m an engineer. And so as an engineer, your natural inclination is to talk about like technology. You know, I’ve been messing with hardware, software, infrastructure, cloud, security, public cloud migrations, you know, talk about any of the stuff that we can talk about and pretty, pretty, you know, eye level as well as low levels for the depth.

Robby Gulri (37:03.194)
so I love to talk about this stuff. I love to do this stuff. I’m a hands on keyboard kind of guy, at least I used to be, where I could jump into a Unix terminal and start doing stuff. Like I love talking about that stuff. But unfortunately, what we’ve learned sometimes through hard lessons is that’s not really what, you know, kind of moves the needle. You know, business owners, operators, leaders, don’t, don’t, they need the technology, but they really need to understand what the technology can do for them. You know, ultimately can it solve business challenges? You know, number one.

Can it, can it, I don’t know, attract more customers? it, can it engage customers in ways that they’ve never been engaging for? Can it help deal with remote work? You know, these are just examples of how technology can be used to help elevate a business, right? So to me, it’s the, it’s the journey of all of us. It’s the collective journey of going from what I call skew to solution. That’s my book ultimately is.

is to shift from skew to solutions. The word skew or the term skew implies that it’s transactional, right? Hey, I’m going to, I got a, I got a widget, Shawn, I’m going to sell it to you. I’m going to quote it to you. And you basically paid me X dollars for it. And guess what? Now you have it. You can use it however you want. That’s not how businesses buy today. So what they basically have shifted to any industry, we can talk about, you know, the consumer lifestyle industry. We can talk about travel and entertainment. We can talk about technology.

They’ve all shifted to and shifting to, you know, being an experienced company, a solution company, an outcome company versus a, you know, a widget or a skew or a transactional company. And that’s kind of the, idea with IT. And that’s why I still like more and more IT is, you know, making that those fundamental shifts, you know, as much as I can.

Yeah, absolutely. know, I’ve said this a couple of times on this podcast, and we’ve only recorded a handful of episodes, so I feel like this is going to be one of those things that it’s my catchphrase. But, know, I’m a hammer and everything’s a nail, so I look at everything through the lens of brand and marketing, And I think that technology, when implemented strategically, can give the company an advantage in the marketplace. It can give them competitive differentiation. It can give them…

Shawn Cordner (39:17.838)
an opportunity to message in a way that their competitors aren’t able to. It’s going to allow them to occupy a space in the marketplace that maybe their competitors can’t. Technology can enable incredible customer experience, which can be a pillar of your brand. There’s so many different ways that technology can not only be additive to the enterprise in the sense of it creates efficiencies or it creates more cost effective.

you know, a way to get things done, but it can actually be transformative in how the company goes to market and how they create products and align the value of their products with the need for the market. when I do branding projects, when I was doing Rapid Scales brand, what I think about is I think about a Venn diagram. And one of those circles is what the market demands and the customers need the most. The other circle is what the competition fails to deliver.

And the third circle is what your company does best. And it’s that little overlap in the center that creates the foundation of your brand strategy. just talk to me about how technology enables that. And maybe let’s make the shift back to AI in that conversation.

Yeah, I can’t, I couldn’t have said it better. I spending a little bit of time in marketing, was in a marketing leadership position. Brand wasn’t like, colors and logos and website, or there’s an element of that, that is true. ultimately it’s a promise that for making to somebody. I promised to basically deal with this value to you, with these solutions to be able to, you know, to solve these interesting, know, challenging problems. That’s really the brand, more of a promise than anything else. But I think technology is,

God, it’s a tool and depending on your lens and depending on how you look at it, it’s the tools from all different walks of life, right? If you’re talking about true engagement, well, there’s a whole slew of technology that can be used to basically make that happen and measure the success for that. If we’re talking about, I don’t know, implementation or migration of applications into some kind of public cloud, there’s a whole slew of technology that can basically be enabled to do that. So, this depends on what lens you’re coming from, but I think more and more,

Robby Gulri (41:29.896)
Leaders aren’t as fearful. I’m not talking about to icy loonies. I’m talking about non-technology leaders. They’re not as fearful of technology. They know that technology is an enabler. They just need to figure out and team up with partners like RapidScale that can help them, you know, ultimately elevate their business. And so that’s why I do what I do. Shawn is I get excited about, you know, ultimately I’m a business person more than I am a technology person as much as I love technology, but it’s the intersection is the convergence of technology and business outcomes ultimately that we’re trying to.

It’s trying to get everybody to think about.

Yeah, absolutely. think from a process perspective, technology leader should be, and most are, right, but I think it’s worth saying, they should be having conversations with the stakeholders in the organization, with the different departments, and understanding where are gaps and where are opportunities. And then they can be the glue that kind of brings together the business goals with the technology goals, right? And I think…

I’m going to keep trying to pull you back into this, but I think that AI right now is the glue because you’ve got the boards saying we need to get into AI. We don’t know what that means. We don’t know what the impact is going to be, but we know we need to get into AI. And you’ve got business leaders saying we need to leverage AI for this and we need to leverage AI for that. And there’s the promise of value of AI, which is what we kind of started with here.

To me, AI is the common language right now that everybody is speaking and it’s really kind of the nexus of the business conversation and the technology conversation.

Robby Gulri (43:07.486)
I, I, I love this AI sort of, motion, you know, velocity, trajectory color, whatever you want that we as a industry, as a, as a country, as a, as a world are having, because honestly, if you’re an advisor, out there and if you’re an IT leader out there, it actually forces you to immerse yourself in the difference. You know, it’s no longer.

Hey, I’m not just talking about network connectivity. I’m not just talking about security solutions. I’m talking about a fundamental shift in the fundamental way to basically change your business to do things that otherwise have not been done before leveraging the data that we talked about for now 30, 40 minutes that you have along with the data that you all have. I’ll give a couple of examples and I think that’ll, that’ll finish this off. I think it’ll.

hopefully resonate with the audience is, you know, we did, we did work for an automotive services company, Shawn, and they said, look, today, my challenge is pretty simple. Actually, they’re, they’re, they’re, so let’s step back. Automotive services company, they come to you. So if you’re, if you’ve got your car at the office, they’ll come to you. If you’ve got your car in your garage at home, they’ll come to you. A repairman is sent for repairs, oil changes, et cetera. You get the, you get the idea. And so the client.

this client of ours basically wanted to make sure that those technicians that were sent were following certain security, and compliance, driverless rights and regulatory stuff. so. So you would ask like, how does AI have to do with it? pictures are taken when a repair person is on site, before AI, they were literally, they had humans behind the scenes in the back office, literally doing nothing but categorization. They were saying, okay, was safety.

followed here, various criteria, was all that followed or was it not followed? And you had humans basically having to make that determination by looking at literally JPEGs or PNG files on a daily basis. So imagine if AI could do this, if a machine could be taught to scan the computer vision of the actual image and be able to make that determination. That’s exactly what we did. So instead of what they called human labelers, easy for me to say, we basically created a

Robby Gulri (45:26.548)
a tool on something called AWS recognition. It’s a microservice that exists on AWS. So ultimately we have now the ability to evaluate images and detect instances when safety mat was used and safety protocols were followed. And the machine can do that. It’s literally almost at a hundred percent. So, so the beauty is that the organization sees immediate value in ROI because they don’t have to necessarily have to allocate those labor resources.

and they can reallocate it into something else.

I that example because it hits on some of the key themes of the promise of AI. I jotted some notes. It’s better use of humans. Those resources could be deployed to something else. Those resources also probably get burned out pretty quickly by staring at these pictures all day long. Employee satisfaction will increase if those people are deployed to different roles.

We’ve got better compliance. know that AI is way better at looking at pictures and discerning things from those pictures than humans are, especially on a repetitive basis. We see it with healthcare and radiology. And then cost effectiveness. It’s much more cost effective to have the AI running rather than a group of humans that you have to pay hourly to sit there and do that work.

And the whole agreed to 100 % across the board. That whole computer image use case, the reason I bring it up is it has applicability. The technology and the process and the methodology has applicability across a multitude of industries. You mentioned healthcare. You’re absolutely right. My cousin’s a radiologist. So he tells me, he’s like, man, there’s a lot of innovation happening around AI and scanning the radiologists. Not for, again,

Robby Gulri (47:15.266)
Today we’re not relying solely on the machine to make a decision or to scan for cancer cells or whatever, or a broken, you know, tubule or whatever. Ultimately, it’s about augmenting, you know, human capabilities and experience along with what the machine can basically balance.

Absolutely. that definitely has applicability. You know what it is? think a lot of times there’s a failure of creativity. We can understand what the value of AI is. We can understand what the capabilities of AI is, but it takes a creative mind to be able to look at the business and say, how can this apply within our operations and how can we leverage it to create an outcome that’s better than what we’re doing today?

And that’s why, we, we, we, we talked a little bit about the whole notion of personalization, prediction segmentation. You know, I’ve looked at, hundreds of use cases. They, they mostly fall into those three categories. You know, prediction is what will happen. Can I predict something that’s going to happen before it does? That gives me a lot of, proactive, and, and, and insights into my business and a competitive differentiator around that business. Can I personalize? Can I…

Can I show or do something for some user differently than another user? That’s ultimately what we’re talking about when it comes to personalization. And then segmentation is, who are the similar groups? Like, you know, A, group B, group C. If I know that, then I can do geo-segmentation. I can do population segmentation. I can do workforce segmentation. I can do all kinds of interesting things and turn it basically target, right? So I always go back to the prediction personalization segmentation. And when I explain AI to people and that, they get that because it’s not about

math anymore. And it’s not about cloud and security. It’s about real human problems that we’ve been frankly trying to solve for a long, long time. We just happen to have some tools that we didn’t have a few years ago.

Shawn Cordner (49:13.568)
Yeah. You were talking about logistics and, are you familiar with the traveling salesman problem?

I have read about the traveling salesman problem. Remind me again of what you’re referring.

think basically what it is is a salesman or saleswoman, they’ve got a series of appointments that they have to make that day, right? have to go on site and visit people. So it’s like, how do you construct that route? What order do you do it in to be most effective and efficient? And that was a longstanding problem. So I’m putting you on the spot here, but I’m wondering, has AI solved that problem?

AI has gotten close to solving that problem. God, that reminds me way back in the day when I thought I was really cool. So I was 19 years old. was doing an internship with Georgia Power Company in Atlanta, Georgia. And I even had a company car. I thought I was the coolest kid in town. And I had one of those bag phones. Remember those back phones was three wide back phones. Yeah. And my job was to go around basically north Georgia, all over basically above making Georgia.

and do energy audits in larger commercial malls, industrial plants, cetera. And I remember having to take out those old map books and try to figure out, okay, I’m going to go from here to here to here. I’ve got seven destinations to hit today. How do I do that in a maximum way? Man, I can feed that into AI all day long today. mean, Google maps does a great job of that. Map quest, remember map quest? You know, it was kind of the first iteration of that, but you know, the AI…

Robby Gulri (50:42.882)
The AI is more real time. It’ll feed in real time impact of traffic, weather, road closures. Because again, this is actually a really good analogy. We’re talking about data that’s not static. It’s constantly evolving and moving and dynamic. know, traffic changes on a real time basis. Road closure, maybe not so much so. Weather changes on a real time basis. So all those impact, you know, the travel and salesman problem. And yeah, AI is definitely…

Is it sold at 100 %? I think we’re at 98.8%.

Yeah, there’s even factors like how many right turns are you making versus left turns, right? there’s all these complicated factors that go into it. And what made me think of it was the example you were giving in logistics where I forget exactly what you said, but it sounded like the company was trying to figure out how to navigate going into ports and taking into consideration all the weather and the currents and just the real time traffic that’s happening at the port, right?

So maybe talk a little bit about that use case if you can.

Yeah, yeah, that’s a good one. I love talking about that one. I didn’t know what a Denbridge fee was until I got involved, you know, in understanding this particular use case, right? A Denbridge fee, by the way, for those of you who may not know who’s listening to this and watching this is an hourly charge that every ship basically gets charged on every big dock or every big port in America. New York is different than Baltimore. Baltimore is a little bit different than Savannah, but ultimately it’s upwards of $60,000 an hour. That’s a lot of money, right?

Robby Gulri (52:17.494)
So it’s it’s the company, the logistics company. That’s what we did was we said, all right, can we predict when a ship is going to arrive before it does? Because guess what? If I can be very accurate without prediction, then I can be much more proactive about shipping, sending my offloaders and my stuff to take the products off the ship and the ship can basically head on its way. Well, in order to solve for that problem, you got to look at a bunch of different data sources. And so that’s exactly what we did is we helped that organization define the problem. one, number two is to be able to then,

you know, define the data sources and then be able to build it out, deploy this model within AWS and then create a, you know, very simple, you know, sort of dashboards for them to actually see, you know, what’s, what’s happening in real time. so, immediate ROI, but again, beautiful use case around how AI can be used to streamline your, you know, your business in, whether you’re talking about logistics or supply chain or trucking or fleet management, there’s a lot of use or potential applicability, you know, to that one.

I have to know also what did you do with the lobster company?

Lobster company data is, data was everywhere, but data was sitting in files, literally files, physical files. So the first thing we did was get their data strategy straight. Let’s get it in a place where we can basically do some insights and analysis from. But ultimately it was a supply demand problem. know, companies have been around for a long time and they’re mechanically or manually try to do price fluctuations based on what they anticipate the supply to be.

to what the demand is going to be. So imagine it’s a fairly simple exercise, but ultimately it’s leveraging the data that we have, historical data, real-time data, be able to give some tools to the company to be able to do a much better and much more accurate job of, know, they’ve got certain margin thresholds they want to follow. They want to basically have certain level of profitability. So if I know my supply demand curves a little bit more appropriately, guess what? I’m a little bit more efficiently approaching that.

Shawn Cordner (54:14.542)
Yeah, that’s probably something that’s applicable to everybody in one sense, but given what’s happening with kind of the uncertainty in global economic markets and particularly, know, tariff prices and how it’s changing and, you know, supply chains right now probably have a lot of question marks. Is AI a good use case for figuring out how to time your supply chain?

shipments or purchases. Talk to me about how that could fit into creating stability in your supply chain and also price stability.

Yeah, I think, you we’ve generalized, probably should have started with this, you know, when we started this podcast, but we’ve generalized like AI with machine learning, machine learning with AI. lot of these terms are used sort of interchangeably. lot of what we’re talking about, frankly, is machine learning, right? Is, is, I, can I use historical data? Can I use, you know, data repository to be able to like, train the machines to be able to kind of…

you know, do from inference, you know, kind of provide some trajectory, provide some prediction. So a lot of what we were talking about in these use cases is frankly more machine learning. A lot of what AI is, is really, we kind of, it’s more generative, starting to be more generative AI. We kind of experienced this when we use, I don’t know, chat GPT or perplexity is the AI is using some statistical modeling behind the scenes to be able to hopefully predict what it is that you want based on the query that you asked. And it’s generating, hence the word generative.

something that didn’t exist before. Right. So that’s kind of the evolution over the last few years. Man, the use case that you’re describing in terms of supply chain management and logistical efficiency and optimization. You know, I think the various manufacturing industries have been doing for trying to do for, for, you know, 30, 40, 50 years, the notion of AI and a, for really the, the, for a machine to be able to act somewhat like a human.

Robby Gulri (56:19.02)
Hasn’t been around since probably 1956, right? The Turing test was one of the first sort of examples of that. And so, you know, what, what’s changed now and why we’re even, you know, talking about these things now is the compute, the modeling, the data sets, the amount of data that’s available to us to be able to find, you know, these patterns and do some of these inferencing and, and, and train, you know, machines to be able to predict some things that otherwise they can’t.

we’ve got this inflection point going on. And that’s honestly why, you we’re seeing this, this explosion of these AI and machine learning oriented conversations.

Yeah, you mentioned manufacturing as being a key vertical there in logistics, but I would imagine healthcare construction, right? Those are two verticals that rely heavily on supplies and at least I don’t know about construction, but at least in the case of healthcare, minimal margins, right? They’re working on razor thin margins to begin with. So I can imagine a world where, you know, being able to ride the supply chain fluctuations and that arbitrage could have an impact.

Yeah. A hundred percent. I don’t know if you’ve talked a lot about supply chain and logistics and manufacturing. That’s exactly right. mean, the, the, the, the amount of data that’s available now and continues to sort of evolve and change. And the amount of data that we’re creating as a, as a society in one day is equivalent to all of the data that ever exists between the beginning of time. So like 1952 or something, you know, crazy, crazy number that I read recently. And so it’s the data that is.

leading to this revolution or evolution, evolution first, then revolution. But yeah, ultimately that’s it. That’s it, Shawn.

Shawn Cordner (57:58.926)
You talked about generative AI and you kind of explained the difference between that and machine learning. I read recently a study from MIT, it was a report that they put out, said 90 % of all gen AI pilots fail. And they have some reasons why. Pilot to scale gap, governance and compliance, cultural and change management factors, budget efficiency.

So I’m wondering, from your perspective, what have you seen from Gen.ai pilots and what’s some advice that you would give to people to ensure that they’re the 5 % and not the 95%. And really, is it that bad if your pilot fails? Maybe it’s a learning experience in your AI journey.

Yeah. I read the same study. It’s actually fascinating. So those of you listening to this, you know, if you go and type in, think the official name of it was the Gen AI Divide MIT. So do a Google search, you’ll find the link and you can download it. It’s a free thing. But yeah, they said, I’m trying to remember the exact numbers, know, don’t, don’t, I’m not in stone. think 40 or 50 billion dollar enterprise investment into generative AI. What they found in the study and only 5 % saw.

you know, some fundamental sort of like value or ROI that was measurable in P &L impact. That’s a pretty, pretty alarming, right? My advice would be again, going back to business outcomes, let’s start with business outcomes first. You know, if you’ve got something in mind that you’re trying to accomplish, man, I’m telling you that’s going to change everything. A lot of organizations don’t know what they’re getting into and they’re like, well, we want to the power of AI. And like, man, I don’t really know what that means. Let’s dig in. Let’s dive in.

So I think that’s, that’s number one is having a business outcome as we wrap this up. The other thing is, is, you know, making sure you’ve got the right partnerships in place. There’s no company that I’ve run into that can do it themselves. Sure. They may have the greatest IT talent. They may have the greatest public cloud talent, but they don’t have data scientists. Usually they don’t have, you know, data engineers that don’t have analytics people all under one roof. There may be a handful of companies in the world that do, but I doubt it. So taking advantage of partnerships, I think is super important.

Robby Gulri (01:00:10.726)
And I think the report actually alluded to that. So I would recommend, you know, go out, you know, find companies like RapidScale. It’s not us and somebody else like us that basically can help you. It’s for sure. And I think the other, like kind of the third element of kind of my thought or my advice, you know, to the, to the organizations would be to, to look at, guess, look at, you know, it’s, it’s sexy to talk about like the stuff that is more impactful to customers, to

to something that’s visible, right? Transparent, tangible. But there’s so much opportunity for ROI generation in the back office. Efficiencies, we talked about so many use cases today where it’s back office functions, frankly. You’re automating, you’re optimizing, you’re reallocating labor into doing other things. That’s immediate ROI there. So I think that’s also what the study alluded to is like, there’s some back office stuff that you guys should be paying attention to that is immediately actionable and recoverable.

So let’s focus in on that as well.

Absolutely. And the stat is external partnerships see twice the success rate of internal build. So I want to echo your sentiment to get help if you’re dipping your toes in AI and you’re not quite sure where to go. Or even if you are, you’re going to do better if you’ve got a partner like Amplix and RapidScale helping you out. Robby, you’re our first guest that had a book to plug and what’s a podcast without a book plug. So.

There you go. There you go. Wow.

Shawn Cordner (01:01:38.894)
Maybe just give us the name again and tell us when it’s coming out and where some people find you

Man, will connect with on LinkedIn, obviously, at RapidScale. So, Robby.Dulry at rapidscale.net easiest way to email me, connect with me on LinkedIn. Man, I wish I had a date for you, man. I am so fired up about it because I’ve been working on it for two years. And it’s hard. It’s one of the hardest things I’ve ever done is to like…

put my thoughts in a structured way on paper. And it’s going to be, I still don’t even have actually an official title, but skewed a solution. And my goal is by the end of the year to have it available on all the typical platforms, Amazon, know, Apple store, et cetera, et cetera.

Awesome. I feel like the podcast thing to say is we’ll have you on when the book comes out. appreciate your time and expertise, man. Thanks so much.

Thank you, Shawn, for organizing this. was really fun to lead this channel. Looking forward to doing it again.