In this insightful conversation, host Michael Bernzweig speaks with Eric Garrison of WTE Solutions and discusses the challenges and solutions in digital transformation for businesses, focusing on data integration and operational efficiency. He emphasizes the importance of creating a 'Lake House' for data, enabling seamless communication between systems and reducing reliance on manual data entry. Garrison also highlights the role of AI in enhancing customer service and operational processes, providing examples from various industries.
In this insightful conversation, host Michael Bernzweig speaks with Eric Garrison of WTE Solutions and discusses the challenges and solutions in digital transformation for businesses, focusing on data integration and operational efficiency. Garrison emphasizes the importance of creating a 'Lake House' for data, enabling seamless communication between systems and reducing reliance on manual data entry. Garrison also highlights the role of AI in enhancing customer service and operational processes, providing examples from various industries.
Takeaways
Sound bites
Chapters
Michael Bernzweig (00:06.396)
boy, that was such a good session. We've got another amazing one for you coming right now. We have an amazing guest, Eric Garrison, who leverages his expertise in SaaS and digital strategy to power WTE solutions transformation programs. His background enables clients like Carfax, Cleveland Clinic, and healthcare companies to achieve digital innovation, drive operational efficiency, and elevate customer experiences. Let's get right into it. Eric, take it away.
Eric Garrison (00:31.946)
Thank you so much for having me today.Today, what we're going to talk about is taking those islands of data. So almost every company today is putting in the who's who of the Gartner top applications. And what we're seeing happen in the marketplace. So we're in with a lot of different companies. Most of them are private equity backed companies and working with the C-suite. So I work as a fractional CTO a lot of times for a company.
and do this where we're coming in and looking at the different elements that are inside of a business and what type of pieces of technology are not talking to each other. One thing that I have noticed is that lot of CEOs are going, I have this great bit of KPI data. Let me pull it up on a dashboard, scroll through. This is the one number I need on their phone. And then they're going back and looking at the bazillion spreadsheets. And that is something we're seeing a lot is that the
CEO is having to log into multiple systems the data that he really needs or she needs to drive their business is Not in a single pane of glass. It's in multiple systems and what we do is we like to take our our clients that we work with and take their data and build a lake house so you've heard about data lakes data lakes are You know a lot of providers out there you take that unstructured data you throw it in there What I'd lake house is it's where you're taking that data structured and getting your systems?
to talk to each other and get you down to a single pane of glass and have the data highly structured so that you can take actionable actions off of that data, but also have it so that it is a system where you've got system A, B, and C all talking to each other and using the same numbers and using the same identifiers between them. So what we see happening with these various islands
is that the CRMs today, let's talk about Salesforce, you that's the 800 pound gorilla of the space. So much of the data is being manually put into it. So with this Lake House type concept that we look at, we want to take and eliminate the human wherever possible. So we have used systems like a dial pad, a five nine, they've got really good AI technology about them. So in the case of like a Salesforce, instead of having somebody
Eric Garrison (02:54.024)
wrap up the call inside of Salesforce and go to the next caller in the call center, we will take the data from the call, pull a CSAT score, pull the transcript, even through this multilingual, write that down to the record, that CRM record inside of Salesforce so that the person is not having to write up their notes, the system's writing up the notes. We also will have other third party systems, let's say the ERP feed the information in that the agent might need. So,
A lot of what I see today and what is needed when you've got a call center rep that's getting on the phone with a person, of course, we're trying to use AI wherever possible to make it easy for the end user, make it so that they're not sitting there hitting zero, zero, zero, and then getting somebody that's not able to take care of them. We want to empower that agent to be able to quickly resolve the issue that is resulting in that human calling into a business today. So we're going to then...
take a bunch of that data from the CRM and put that into the screen pop. A lot of times that involves pulling from two, three systems. We're working in the healthcare space. So a lot of times what we'll do there is pull in from the EHR system what their next coming appointments were, who they last saw. So when that screen pop pops up for the person, they're able to get it. That brings great KPI because that brings your call time down. That's what a CEO, CFO wants to do. We got to come up with AI solutions to actually move the financial needle.
So the best way to get a budget approved in a organization today is also come up with those ways to save costs, save time. These departments, know, rain their own EHR system in the clinical side of a system, then you rain NetSuite on the financial side, getting those systems to talk and use the same information and track the patients or track the clients in that journey.
is going to be very important to make sure that, let's say we're dealing with a company like a Glaxo. Well, Glaxo's got five names. There are the Glaxo, GlaxoSmith, GlaxoGSK. You want to have that come where you're dealing with the systems that are talking about that potential partner company and tracking their data under one name. So that's where a data normalization layer is going to be very important. And it helps you when you're building these.
Eric Garrison (05:15.018)
these journeys, whether that's a customer journey, a patient journey, whatever those critical decisions are, we used to call them funnels, but let's face it, today's funnels look more like a ball of spaghetti that is out there because patients are coming through multiple channels, clients are coming through multiple channels, so being able to track them is very important. And in organizations today, what we hear a lot is we're very dependent on Excel.
have these people in every office, every client we talk to, every CEO, they've got their person who is in charge of doing the pivot tables and the VLOOKUPs. Hopefully it's not the CEO. I have had a couple of occasions where CEOs are asking me questions going, what's the best way to build this really complex VLOOKUP to my pivot table and get it to refresh? My SharePoint file is not working right. That is where we need to really break that cycle and start putting in actionable data.
feeding into platforms, not trying to slice and dice it and keep. I have a number of clients who live and die by their KPIs in their Excel spreadsheets, which makes the time that it comes to put the board deck together really hard. So then you're having to find, you know, whether it's two, three months worth of Excel files, pull those all together, as well as tracking budget items. So that's where a product like a Data Rails,
which is very Excel focused. It is part data warehouse, part data lake on the back end, can tie systems together. That's one you may want to take a look at that is, you know, I hate to throw another tool to fix tool problems, but that is a place where having a data repository that can pull these multiple things together and then feed that data into Excel intelligently. So what it does, takes an Excel sheet that you've been using forever and makes it intelligent and makes it pull from their cloud.
the real data on a refresh. So we've had a couple of clients have taken and they had a person that was spending eight hours a day just updating the Excel spreadsheet every Monday to do what was last week to get the KPI numbers in for management. So this is able to save them a lot of time. And also that person in case couldn't go on vacation. So we hear a lot of, know, what do we do when Sarah who is our
Eric Garrison (07:38.635)
Person or Josh who's our person who is in charge of all of these Excel spreadsheets and all of our KPI numbers? We have to skip a month of data here Analysis because that person took a week vacation So that's something we want to make sure that your organization is able to avoid when you are bringing these data sources together It makes it much easier So, know, this is where you're gonna want to make sure you take your main pivotal ones
you your NetSuite, your CRM, either HubSpot, Salesforce, whatever, as well as your communication channels. I like to take and pull data to and from in a bidirectional fashion from a product like Microsoft Teams, Slack can work as well, because you can now have those tools interact and be a conduit of where the data is coming together in a collaborative fashion.
very exciting stuff and Copilot can work for this as a good AI intermediary to it. But then what you want to do is have a central repository that we're calling the Lake House. So this could be a SQL Server. This could be a relational database out on any of the cloud providers. This could be out on Azure. There's a lot of different options that you can do here. We like the Microsoft Stack, so we're a little preference with the Microsoft Stack.
Power BI is a great tool to get these dashboards into. A lot of cases, what we have found though, is there are limitations to some of these and the time it takes to get into Power BI. And a lot of people then still take the route of building your own custom dashboard system. So when you want to have a bespoke system that gives you exactly what you want, don't be locked in to say, it all has to be Power BI, because sometimes these off the shelf tools like that are not the right tool.
and an integrator or somebody could create a dashboard system that feeds it into Excel if that is what's going to be better for your system or it could be a web interface. What you want to do is have it so that your subject matter experts on that data feels like it's their data being represented in there. They can trust those numbers. Trust is a big thing when pulling together your KPIs. What you don't want to do is have a CEO that doesn't trust and have the SMEs underneath of that CEO.
Eric Garrison (10:02.661)
not trust the data that's going into these systems. So that's where cleaning, deduplicating, validating all of this data is very important. Make sure that when we're aggregating the data, if we're saying that we are serving this number of clients or patients, that this system has the same number, that is a very difficult thing. That is a very hard process to do, but it is very worthwhile to make sure that all of your BI and systems are
on the same page, all the same numbers, all the same outcomes from it. So that's a very important piece to it. The other thing to watch is you're going to find how much chaos there is. Every finance department I talk to, I'm really shocked at how long it takes to close out the books. So we've had one client that we're working with that said that the CFO told us on the phone call recently.
Eight days is what it takes to close out the books and that's down from 12. Luckily, they have put in some newer technology and are able to get that down. But his goal is to get it down to two or three. Another company is running at 20. You don't want to be sitting in that space. And what you also don't want to do is also from a C-suite have to be able to go through all these different iterations and not be able to get your prior month data for 20 days. That's a killer because
You're driving the car on a road that has potholes, but you don't know whether those potholes exist in what lane and how fast you're to be able to go and whether you need to pull over and put new tires on the car. It's a very bad situation when it takes too long to pull all that together. That's where using things like some of the off the shelf products that have smart connectors to them. So, you know, like a net suite is able to talk via Zapier or Make and you know, some of these platforms.
you can take a shortcut and bring some of these together. No code frameworks are really good, but here's my suggestion. No code frameworks are not a replacement for a developer. You're going to be best to use a no code framework with developers to make your developers better and to bring your platforms together. AI co-pilot, so not just Microsoft co-pilot, but bringing these systems into your platforms. So I know NetSuite's putting it in and we're...
Eric Garrison (12:25.854)
They're able to help you close your month. They're also able to put things in there to help you analyze some data. This is just in its infancy. So think about this way. You want to start looking at how we can get this structured data in a local LLM or dedicated LLM that is able to have a lot of your company data to be able to analyze and look for where things are going and also what is happening.
in your day-to-day business to be able to have that business intelligence, not just in a BI dashboard, you want that BI data and intelligence feeding into these other applications. Let me give you a case here of a, you know, everybody likes a chat bot. So here, think about this from a chat bot. Let's say that you are a big high-rise apartment building in a high-priced city. You've got 40 stories worth of residents.
A resident fills out a ticket, they may be asking, can I get Uber Eats delivered? That's not a reason for AI, that's just decision tree. Now let's say that somebody fills out a ticket inside the chat bot and says that, hey, I have a water leak. That is a case where you want to use machine learning AI to diagnose, ask a few other follow-up questions. But also as you're doing that, hit your data warehouse and hit this data lake to say that this person, they put in their information, their phone number, something, we know that this is
Susan on floor 40 in the penthouse. We need to make sure that we get there and get out there and here is how we can drive this ticket and alert everybody send out SMS. That is a great use of using chatbot AI technology, decision tree technology, machine learning altogether and bring it full circle because what you're looking to do is have these systems that have got the intelligence to drive your business from the CEO's suite as well as CFO.
and all the way down the pipe. you know, what I'd like to tell people is that when you are able to bring these type systems together and then show that actionable data of how that ticket, even in that high rise building or in healthcare, how much money you can save by deploying these, it makes it even better. It makes getting budget for AI and BI projects a lot easier. But when you have a single pane of glass that you are driving your business, it's what a company like Cisco Meraki has driven their whole business off of.
Eric Garrison (14:50.524)
is that single pane of glass. Do the same thing for your business, for the C-suite to be able to have zero Excel files driving the data decision, because that a lot of times is flat data and historical data. Let's get it down to where we are driving the data off of real-time actionable data, but also predicting where we're going before we get there. And that is where AI assistants can bring that level of predictability and say, hey, based on where we are today,
How does our quarter look coming up?
Eric Garrison (15:26.43)
That'll be my wrap.
Michael Bernzweig (15:31.992)
Thank you so much for that wonderful presentation. I know a lot of people had some questions coming in. And guys, remember, we're going to have a Q &A session in just a minute. So he's going to answer all those questions. And we'll see you guys in the next session.