Swimming With Sharks: Enterprise GenAI Unplugged - S3 Episode 4: Rob Ryan

In this episode of Swimming with Sharks, Kevin Dean, CEO of ManoByte, is joined by Rob Ryan, Head of Thought Leadership at WorkGrid Software, for an in-depth discussion on the role of conversational AI in streamlining employee workflows and improving internal communication. Rob shares his insights on organizational change, the evolution of employee experience platforms, and the importance of aligning AI projects with business goals.

Introduction: Kevin introduces Rob Ryan, who has extensive experience in organizational strategy and technology deployment. Rob provides background on WorkGrid Software and its mission to reduce digital friction in the workplace.

Interview Highlights:

• Conversational AI as a Digital Assistant: Rob explains how WorkGrid leverages NLP and machine learning to create digital assistants that streamline employees’ day-to-day tasks. He discusses how conversational AI can reduce digital distractions by integrating with existing tech stacks and bringing actionable insights directly to employees.
• Getting Started with AI Projects: The conversation explores practical steps for organizations to initiate AI projects, emphasizing the importance of quick wins and internal use cases. Rob shares advice on piloting AI initiatives, mitigating risks, and developing a governance strategy to ensure AI aligns with business objectives.
• Transforming Employee Experience with AI: Rob highlights how conversational AI can enhance the employee experience by reducing app hopping and context switching, ultimately fostering a more productive and engaging work environment. He emphasizes that AI should assist, not replace, the human workforce.

Key Takeaways:

• Building an AI Strategy: Rob advises organizations to start with low-risk, high-value AI use cases that support internal goals and gradually expand based on proven results.
• Importance of Governance: A strong governance framework is essential for AI projects to ensure that they are sustainable and that technology is used ethically and effectively.
• Prioritizing Employee Experience: By integrating AI into the flow of daily tasks, companies can improve employee engagement and reduce friction, allowing employees to focus on high-value work.

Transcript

0:02

Hey there and welcome to Swimming with Sharks, a deep dive into the world of AI for the enterprise. I'm your host, Kevin Dean. I'm the CEO of Manavite and I'm thrilled to have you join us for this exciting journey through the dynamic world of AI use cases, platforms, tools, and best practices. I am super excited to have with me today Rob Ryan from Work Grid. Rob, it is great having you on the show today. Thanks, Kevin. Really appreciate you having us on today. Yeah, so you and I have had lots of great conversations. Just for a moment, tell the audience a little bit about yourself. Give them some background. Yeah, absolutely. So Rob Ryan, I'm the head of thought leadership and business value at WorkRid Software. I'm also part of the leadership team. A little bit about me, I've spent my career in and around strategy, technology, and organizational change management. My early career was the first decade was spent at a little company called State Street. where I focused in on strategy and deployment of Enterprise 2.0 technology. So if we go back to 2008 or so, when all that social collaboration tools were new, I was part of a very small Tiger team helping to introduce those solutions. And after that, I pretty soon jumped over to the vendor side, working at a number of intranet vendors, and jumped over to the conversational AI side about three years ago here at WorkRidin. Wow, that's a pretty cool history. I think we're going to dive into some of that as we have some of our conversations. But tell me a little bit about WorkGrid. Absolutely. So WorkRid is a conversational AI platform that effectively cuts through the clutter of an employee's day, streamlining workflows to nos, to dos. And it does this by leveraging NLP, natural language processing, and machine learning. So you can think of us as a co-pilot or a digital assistant to the employee's day that can be embedded into any portal, browser, teams, really follows the employee. wherever they happen to be. And the best part is it's able to connect to all the systems and applications that you have within your tech stack. And why is that a value? Because if we're able to really bring information, action, and utility wherever the employee is, we're reducing that digital friction. We're reducing that constant app hopping and distractions because most employees, they'll use anywhere between 25 and 35 applications per day. They'll context switch anywhere between 1,200 to 3,000 times per day for your knowledge workers. And all that creates is a lot of noise, grit, and pain throughout the day. So if we can minimize that, if we can create a better work day for the employees, where we're doing plus one for the organization and for you enjoying your work day that much better. There's a lot to unpack there, but what's interesting to me is there seems to be this connection between your background of organizational development and WorkGrid and the platform. Can you tell me from the lens of your organizational history, how should people be thinking about use cases for conversational AI tools like WorkGrid? Yeah, that's a great point. So one of the challenges that we oftentimes hear, and this is industry-wide, I should say, is that the business side is oftentimes thinking, well, where do we start? Where should we start? And in fact, I was at an event a few weeks back. In fact, you and I were at the same event, so we heard some of the same talk tracks at Forrester. And you heard from folks on the business side. We don't know where to start. What use cases should we begin with, particularly around Gen.AI? And the IT side is saying, well, look, we're mitigating risk. We're having to assess platforms like they're not ready yet. They're trying to size things up. And of course, the analysts have their own story. And so for organizations who are looking to really start, it's looking for those quick wins, those quick trials that can quickly showcase value to the organization. And there's many use cases out there. to get started that are in a safe way that you can quickly showcase business appropriate use. But a lot of that comes down to what is going to be our strategy, what is going to be our governance. This technology, while it may be new for organizations, all technology challenges come down to the same, as you mentioned, organizational change management approach of that it's usually a people problem and not necessarily a technology problem. So the right strategy, the right governance. the right organizational change frameworks, that is what ultimately will make success. Yeah, that is awesome. You know, one of the things that people think about when it comes to AI, GNI, conversational AI, agenda AI, you mentioned it, where do we start? Do you think it's good to start with internally facing applications before you start taking things external and why would you go with one approach over the other? Yeah, it's it's you should take the crawl walk run approach because of risk mitigation. Ultimately, look, we saw it in the headlines, right? There was the Canada Air fiasco where they decided to hook up their chat bot, which started, you know, hallucinating to customers and, you know, basically positioning themselves as I think it was flights that they were able to gain refunds for. There was the Samsung incident that happened last year with the data leak. That should all be trialed and tested, pressure tested in-house. So starting small on use cases within pockets of the organization, piloting, and then justifying that value is really where to begin. And you can begin to actually map and measure out your use cases based upon the risk profile, the impact, who's going to own it, and really decide and define that program strategy that will ultimately be successful. You do hear a pattern sometimes within the industry as a whole where people are just simply, okay, we'll do a co-pilot pilot, right, with Microsoft. And of course you're getting charged $30 per user per month, but so be it. And then people are just, you know, chattering and chatting away to the bot, but there's not necessarily a focus into how we're going to use it. And so there's going to be some messiness there. You could... pick the counterpoint of saying, but we also may find some gold in some of those edge cases. More often than not, you're going to have a lot of noise that someone will have to sift through. But if you pare that down into, we're going to focus in on some of our marketing efforts. We're going to focus in on streamlining sales enablement. We're going to focus on customer support and success to have them transform the way they engage with their customers. Those are going to be wins. So really defining that audience, defining that scope, those use cases, and then being able to justify your value of both the solutions that you're bringing to the table, that's what gets customers ahead. External, that should be when you're further along on your AI adoption, your governance framework is that much more mature and stronger, then you're able to meet those needs externally. That's very, very cool. when you think about like WorkRid, right? And how that it can help improve your internal employee experience. Two questions. Back to how do you get started with a tool like WorkGrid and how long of a implementation initiative is it? Like, you know, I'm hearing things like a lot of these GNI conversational AI projects don't take as long as companies think in order to get them up and running. Give me some insight into that. absolutely. Part of that is, part of that is trying to, it's the misjustification that the conversational AI platform is a chat bot. So part of the market has been sullied by this past of chat bot, if you will. and there's actually some conversational AI vendors who actually position themselves as conversational AI. But if you look under the covers, it's like, wow, this is a PS. This is going to take months, if not a full year, to wire everything together and to quickly see value. The more modern conversational AI platforms, WorkRid for example, we have the platform, the templates, the data sources to quickly stand up those use cases in a matter of hours. Let's say if we want to take time off, sense into pay, we want to chat to particular knowledge sources like ServiceNow, SFDC, Workday, etc. those quick start ways to begin to showcase and trial conversational AI. Where you see the separation is perhaps an organization may have tried out, let's say a legacy chatbot or a legacy interface at the time. And then once they started unpacking it, my God, this is going to take us months to see any form of value. Today, you could see that within weeks, depending on the size and the scope of those use cases. One of our efforts was to really focus in on our builder platform, which allows those organizations to quickly stand up those use cases based upon templates. But also if they have business technologists, so business technologists, really any folks who happen to live inside the business who have a little bit of tech savvy, really drag and drop within a low code, no code interface so that they can create more bespoke aspects of those templates. So take that template, perhaps it's for pay or whatnot. and you want to create more of a nuanced use case for your organization, fantastic. Really it's fields, drag and drop to get yourself started. So that has been a movement certainly within our particular spaces. How do we get folks being successful that much faster? And that has been certainly part of the groundswell of where conversational AI has been moving. So again, I'm still like circling around like a shark. I'm circling around this concept of we've got technology and tools like a work grid that I can technically get up and running in the matter of days or weeks, right? So we've got the tool that can do that. But you said something interesting, like companies are coming from this skepticism. Mm-hmm. Yeah, but we tried this before and it took years how How do you bridge like how do you bridge this? This this gap that we have of I was here before I'm not sure that this can really do this because I've been told this story Once or how do you get over that? Yeah, you have to really show them the way and end up showing them how fast they can quickly begin to realize value within their organization. So we may meet with a client and they have a particular situation. Let's say it's connecting into knowledge sources in relation to their benefits or whatnot. And if we're able to gain access to that, hey, we'll absolutely show you that within just a matter of days. You have some folks who, you know, willing us to give that or will even show you it, you know, using some dummy content, you can quickly see how that becomes realized. And some of it also is within the actual engagement itself where legacy chat box very much focused on more semantic keywords. So, so long as you were saying it the right way, it's going to retrieve a semblance of a result. LLMs have taken that much further because now it's more intuitively understanding the user's intent. and then defining what those next actions may be. So if someone's looking for perhaps something in relation to an eye test or, hey, do I have more leg room as a part of my employee benefits package or my T and E, it will end up understanding the intent of that, who that individual is, and then where we should look for that information. Now, what might it return? Based upon the disambiguation of that particular utterance, it may recall back a summary of that particular policy, the source of that policy itself, and then perhaps even the next step action for that employee to take, understanding they asked this query, what are the parts of that query, and then what does this individual need to make that next best decision, or how can we even take that decision and make it for them for the next steps? Because we know with confidence that that's where they're headed to on their next journey for that information to action. That's awesome, I like that. But let me ask you this. My systems aren't all cleaned up yet. What do I do in those instances? Yeah, that's a great point. have great point. So one of the ways you can really start to get started is by focusing in on what are those key use cases that don't necessarily require wholeheartedly data sets to be cleaned up in terms of knowledge and resources. Or is this now a time for us to focus in on cleaning up those resources, finally cleaning up the house so you can think of it as, you know, effectively like a move. Like if your attic is full of clutter and whatnot, okay. That's where all our knowledge lives. Perhaps we should focus more on the dining room, which is at least partially clean and presentable. So we're going to focus in on pay time off, like queries in relation to a particular vertical, let's say human resources, because they have their house in order in terms of where to point and look to people. Fantastic. But perhaps some of our KBs in relation to our product or whatnot, we have to clean that up. We have to make sure that's the right information before we position. So it can be a bit of this give and take of understanding. Where is the client on their AI journey? And what can be done in the short term to quickly show that success so that they can prove their value? And what's that next phase as a part of their evolution? So being able to quickly deliver those key use cases, at least some of the long pole use cases of conversational AI, and then having really that backlog of, okay, we do need to make the next steps on some of those additional use cases, the more in-depth transformative use cases that you see. So that's one of the challenges that's happening with inside organizations who are right in that middle or mature stage. They've done some of the quick wins in terms of conversational AI, but the transformative nature of AI for the business, being able to, let's say, sense within their value chain, within operations, whatnot, systems testing and producing risk reports as far as triggers, that may not have been captured as of yet. And really, no platform is going to be able to solve for that. That's the brute work of actually getting your house in order to ensure you're ready for that transformative AI and machine learning that's around the corner. You definitely do see some success stories out there of organizations who have done this work. But the key bit is they were doing this work three years ago. It hadn't reached the excitement as JNAI has introduced with consumer grade technology and us being all exposed to JNAI and kids asking to use the chat GPT. Yeah, so what I heard is that you've got to get the house ready before you can throw the party. If you're looking at some of those late stage adoption use cases, absolutely. For conversational AI, you can get start quite quickly. And even for Gen. AI use cases, let's say you want to introduce more of a conversational aspect similar to a chat GPT or entropic experience where you're chatting to LLMs and allowing your employees to get information out of them. Those are very quick to stand out. We're talking a matter of hours for you to begin to trial that. We offer a solution as a part of our templates called genies. And this is for those organizations who are really trying to dip their toes in the water. And what a genie does is it wraps a gen AI use cases defined by what you determine business appropriate. So for example, you're going to have a genie, which is inside the conversational interface that is targeted towards those within marketing. you're only going to allow for what you deem fit. All right, we're going to allow for email rewrites. We're going to allow for blog generation, perhaps town hall meeting notes or town hall sessions, and then tone that we want to define. Now, why is that important? Well, one, there's constraints there that can also be tagged to whatever LLM you wish to have, or if you want to leverage the work at LLM that's not tuning and training on your data. That's a key point. A lot of conversational AI platforms out there are training. On the organizational data, we're adapting your information. You should be focused on, OK, is this model truly closed? Are they ensuring that it's not going to be training and tuning? That is definitely key. So there's the guardrails that are in place, both from a use case perspective, is a business appropriate, and a technical perspective that makes IT happy. And why is that important? You can quickly see the value. Organizations who are a bit more mature in their growth, they can quickly start to see, OK, you know, we don't necessarily need the genie. We can chat to the conversational interface much like we would any LLM out there. And then we'll also know what we're looking for delivering those actions and taking those general questions that we have. That's awesome. That is awesome. So this is powerful. And I think that folks listening to this show today, I think they're going to walk away with a lot of key insights on how they can get started right away and add value to their organizations. So in general, tell me, what's exciting you right now about the world of AI? Yeah, the speed that it's moving is incredible. What excites me the most is one of the things that we've been focused on, and you can see it as a part of our marketing and a part of our focus, is really guided attention technology. so think of it as advanced, agentic AI. So that's being able to sense and know what you need in the moment, and then also taking those next steps and actions going forward. So most conversational AI platforms acts like basically a digital assistant today. I type something in, something pops out. One of our value props that we focus in on is really being able to deliver notifications, nudges, things that are important to the employee within the flow of their day. And why is that important? Because it's dampening the digital friction. It's bringing that information so you don't have to go to another application, remember another UI. it's in the flow of your day without it being noisy. That's one bit. The second bit there is really the agentic piece. So being able to take those next steps for that employee so they're not having to continue to jump through hoops. And that's at the heart of what we do is if we can dampen that digital friction, if we can guide that attention, we're going to create that better working day for employees. I love that. So when you take a look around, what worries you about what's happening with AI? Yeah, that's funny. I guess it's more of a, this is a much more personal nature of what worries me with AI. It's the potential lack of creativity and humanity for my kids. So for example, you you and I, we grew up with pen and paper in school, right? Passing your report, you were writing on a piece of paper, passing that in all the way up until college till now you had to type it in a double space. And so you had to learn how to write, how to think, how to structure your thoughts. So my worry in terms of AI ends up coming back at home, which is, OK, I want to make sure my kids at least know how to structure prose, how to be persuasive and not necessarily lean on a bot. that is basically numbers under the covers. Machine learning is statistics when you pull off the mask. So it's a bit of that. In terms of AI as a whole within the workplace, don't fear it replacing humans. It's definitely going to change jobs. It's going to change the way you work. Now, there's also the productivity paradox, which is, All the new solutions and tools that get introduced within an organization, has it actually increased productivity? No. So where does that go? And that's always been a giant conundrum as to where does that time goes? Especially if employees are spending more than half their day, knowledge workers, more than half the day searching for information, having to app hop and go through. It's definitely going to change the nature of the employee with their technology. but it should be one that's done under a humane fashion. Hence one of the reasons why I've stayed and enjoy the conversational AI space. I see the potential there of it truly acting as a Jarvis to Iron Man, as Alfred to Batman, like that assistant that's there when you need it, but isn't necessarily in the way or in your face. So it's the shining light that could ultimately enhance the day, keep you moving. but not end up adding friction. That's my hope. Now, there's all those other struggles in relation to, who's going to win the LLM war because this is really going to be commoditized technology. And then what does that mean in terms of all those underlying models and the ethics and bias behind it because those need to be right sized for an organization, right? You want to make sure that you're getting at least an independent truth to your employees. There's all those concerns and worries. I'm not worried about AGI coming through and the Terminator movies happening as of yet. We'll talk to me in maybe two years. Yeah, that's cool. That's really cool. Great things to think about. So what bit of advice would you give our listeners as they're considering thinking through different AI focused projects? Yeah, I'd say focus first on the low hanging fruit, the use cases that can quickly showcase and drive value for the organization that's at least aligned to your business strategy as a whole. Where we're seeing a number of success stories, particularly around particular functions. you know, customer support service, that's not external, that's really internal. So if I can create better customer support, agents and them being able to find the information faster for folks that they have on the phone respond to them faster with the right tone. Well, that's assisting me throughout my day and I'm creating better value for the organization in their engagement with their customers and clients. So it's looking for what is the nature of the work that we do? How can we quickly begin to showcase quick wins to the organization? determine that business value, and then lean into those use cases and expand from there. So taking the classic land and expand approach is usually a quick win for an organization. The other bit is really putting the resources behind this, not just simply throwing it at a wall, budget clawing, but really, what is going to be our AI strategy for employees? What might be our AI strategy for our engagement with our customer experience? What's it going to be for transforming the way we risk mitigate, find new opportunities, transform the way we work and our systems engage with one another? Those are all key criteria. So having the right governance framework, the right program strategy, and frankly, the right leadership in place to bring this about. I love that. This is powerful advice. Very great, great. tell me when you're not geeking out over all this AI stuff, what do like doing for fun? What do I like doing for fun? Well, when I do have some quick time to myself, I got two little ones, so I don't have much of it. I'm either in the gym, I've been doing CrossFit since 2013, and if I can eke out time, I do some artwork. I was a painting major in a previous life, went to the Massachusetts College of Art and Design many, many, many years ago. And then of course I smartened up and got an MBA when I was working in tech. so you gotta share with me some of your artwork at some point. I wanna see something. Yeah, that's actually one right behind me there, back from, gosh, 2001, 2002. wow, that's yours? Beautiful. Beautiful. Love that. Wow. That's powerful. lets me hide things in certain parts of the home, this being my studio. So yeah, you gotta have your wife talk to my wife and give me a little more freedom. So most folks who listen to the show know that I love music. I love music. So give me one song that I should go and rock out to later on today. gosh. One song to rock out to today. I'll give you two. I'll give you two just because I try to introduce new music to my kids. yeah, so Sinner Man by Nina Simone, that is just pure like energy. That is just, we were listening to that on the way to school today and I'm trying to tell them I got to look. This is one of the best singers of all time. You have to listen to this. And then of course, there's certain things I can't let them listen to. But In the Gym, a great song if you're going for a PR is Hail to the King. So that's another good one. man, this, I love this. I've got some great things that I'm gonna rock out to you today. I love it, love it, love it. Rob, this has been so much fun as always. Every time we get the opportunity to talk, it's always great, always educational. I learned something. This has been fun. So thanks so much. We appreciate you coming on the show today and everybody listening, we appreciate you tuning in. That'll be a wrap for this episode of Swimming with Sharks. We hope that we leave you with some fresh insights that will inspire you for your next AI project. So thanks for tuning in and we'll catch you on the next episode of Swimming with Sharks. Thanks, Kevin.