For this episode of the #HRTechChat, our guest was Melissa Swisher, chief revenue officer of Socrates.ai. To say Socrates.ai is an artificially intelligent chatbot to tie together various domains of the enterprise software ecosystem is probably the most straightforward description. The description implies only a small sliver of Socrates.ai’s potential application. Some of these domains themselves may have chatbots specific to various proprietary software, and Socrates.ai an draw information from it all.
As Melissa elaborated during our chat, Socrates.ai “is an employee experience platform. Think of it as an experience layer” that hews to employees’ preferences. My take is this: Think of it as a prosthetic to facilitate a unified experience in employee self-service as this pertains not just to staff’s employment, but to anything they wish or need to know over the course of a given workday. In my experience, ESS for human capital management specifically offers a rich potential vein of return on investment, and I can only imagine that this ROI grows exponentially when applied to more of the employee experience. Melissa and I delved into some interesting numbers around newfound efficiencies deployments of Socrates.ai have produced.
Naturally, because of the solution’s relations to AI, our conversation expanded beyond Socrates.ai itself. And, as mentioned during the chat, I really wish something like Socrates.ai had been available to me much earlier in my career, when I worked with a team at civil engineering firm to figure out what the company had done, in its past, that was relevant to the many responses to request for proposals we drafted. A prosthetic to identify and pull that information from within the depths of the company’s enterprise would have been nice.
Then again, it was 2003, not 2021 — the age of AI. We’ve come a long way, and I encourage everyone to watch this episode.
Our #HRTechChat Series is also available as a podcast on the following platforms:
- Apple iTunes: https://podcasts.apple.com/us/podcast/3sixty-insights/id1555291436?itsct=podcast_box&itscg=30200
- Spotify: https://open.spotify.com/show/0eZJI1FqM3joDA8uLc05zz
- Stitcher: https://www.stitcher.com/show/3sixty-insights
- SoundCloud: https://soundcloud.com/user-965220588
- YouTube: https://www.youtube.com/channel/UC6aDtrRPYoCIrnkeOYZRpAQ/featured
- iHeartRadio: https://www.iheart.com/podcast/269-3sixty-insights-78293678/
- Google: https://podcasts.google.com/feed/aHR0cHM6Ly9mZWVkcy5zb3VuZGNsb3VkLmNvbS91c2Vycy9zb3VuZGNsb3VkOnVzZXJzOjkyNDY0ODI1OS9zb3VuZHMucnNz
- PlayerFM: https://player.fm/series/3sixty-insights
- Instagram: https://www.instagram.com/3sixtyinsights/
- Amazon Music & Audible: https://music.amazon.com/podcasts/7e7d809b-cb6c-4160-bacf-d0bb90e22f7c
See a service missing that you use? Let our team know by emailing research@3SixtyInsights.com.
Brent Skinner 00:00
Well, hello, everyone, and welcome to the latest episode of HR tech chat. And I’m very happy to have as our guest today, Melissa Swisher who is chief Revenue Officer of Socrates AI, which is very interesting. I think it’s a it’s a gross simplification to say that Socrates AI is a chatbot. I am. And that’s where I’d like to maybe start today. Welcome, Melissa. Very happy to have you. Thank you. Happy to be here. Yeah. Yeah. And, and like I was saying, Socrates AI does. It’s a very interesting tool. And I guess it expresses itself most, you know, most noticeably as a chatbot. But it really is much more. Could you just explain for our viewers just a little bit around skip and sort of a an idea of what Socrates AI does? What kind of issues you solve for this sort of thing?
Melissa Swisher 00:59
Sure, absolutely. So Well, again, I’m thrilled to be here and talk a little bit more about this, and, and some of the challenges with employee experience in optimizing for it. But essentially, what Socrates does in a nutshell, is, if you think about it, the world of work is there’s so many different places where employees can go, right, to get answers to questions, to try to figure out what type of information they’re looking for. And there’s so many different systems that people have to interface with. And what we saw in terms of the problem when we started the company four years ago, was really that, you know, there’s so many places where the people need to go to figure out, okay, where is this policy set? You know, how do I update this information? And so there’s just this system and information overload, it’s all there. But oftentimes people struggle with Well, where do I go for what, and so really, at the core of what Socrates does is, it’s an employee experience platform, think of it as almost a layer, right? An experience layer, that really goes to where employees work, in terms of a technology that they’re already comfortable with, in terms of whether they prefer to interface with a company through a chat portal, if they prefer to use text messaging or collaboration tools, like Microsoft Teams, and really what we focus on is really getting full, single answers versus search results, allowing them to complete any sort of things that they might need to update take action on, that goes across all those different systems and solutions that might live in the enterprise. And so if you think about it, hiding all of the noise of what might need to interface with and making it a very simple, very consumer, like friendly user experience.
Brent Skinner 02:51
Interesting. So if I were to translate this to, you know, sort of HCM parlance would it be would it be fair or accurate to call this? A maybe a prosthetic for Employee Self Service, something like that? Or?
Melissa Swisher 03:13
Absolutely, right. So you know, there’s this whole notion of right empowering people for that self service. But really, what we want optimize is, is to give people access to things whenever they need it. And whether you know, if you’ve ever been through an experience where you’re looking for information, for instance, on SharePoint, it’s a great repository. But if an employee is looking for something, what do you there happens is they get 1000s of results, right? In terms of what the potential answer are, they get none at all. And so our whole idea is to break that down and make it super simple, just in the world of consumer world, right? Like, think of an Amazon experience as a consumer, right? You know, they do a fantastic job with the algorithms behind the scene is if you’re looking for a product, they’ll serve up vendors who supply that product, they’ll give you the reviews, and they’ll give you your information of who’s going to get there soon as what’s the price, you don’t really care about all the sausage making that goes into it, you really want to get your answer, ability to go get what you need, and move on with your day. And that same thought process is what really we saw is the gap that was missing from an employee and the workforce experiences that people have.
Brent Skinner 04:23
One of the things is that just from my background, in a previous life, as an analyst before now, we used to look at the ROI of, of cloud software deployments. And that’s sort of a gross oversimplification. But one thing one thing that this makes me think of is whenever we looked at a deployment where there was a modern Employee Self Service, solution deployed, something that you know that that replaced nothing or replaced some really sort of old, clunky, not really all that useful, you know, 10 year old emss, or whatever, right? We would always see a noticeable gain, aggregate gain and productivity for the, for the, for the workforce, right? Because everybody’s saving maybe five minutes here and there over the course of a day, and the larger your workforce that that really adds up. And, and then when you look at that as a productivity gain, and maybe you, maybe you apply some sort of a correction factor to that just to just to just to be as conservative as possible, right? Because people aren’t spending all of their newfound time on something more productive, especially if it’s very small chunks of time of the course of the day, right. But they are still saving time, and they are being more productive in the totality of things. But with a large workforce, we’d see, you know, we’d sometimes see a productivity gain. I’d have to go back to my old notes from those days, which would, which are not right at hand. But, but I do recall a couple of examples where it was, it did amount to a couple $100,000 worth for very large organizations, it just depends on how many folks you have there that are that touched the new system that that makes things easier for them.
Melissa Swisher 06:33
Yeah, absolutely. And that’s something that we see, you know, with our customer base, and people that we interface with is that, you know, oftentimes, right? What we see happen is it’s a productivity piece. But it’s also, if you take a look at where people get stuck, and if even if you take a look at shared services and call centers, right, oftentimes what people are calling into fall into three categories, right? So it’s the, you know, what is this are the simple repeated questions that come up there, you know, over and over again. So those repeat questions that are pretty simple, but standard FAQs. The second thing is really about these, how do I write? How do I do this? You know, or where do I do that. And then the third piece is really around I need to write. So all of those things indicate an employee’s employees frustration level of trying to figure out almost where they go. And oftentimes the things are so repetitive in terms of what’s coming in. And so really, there’s two parts, right? If you think about that productivity, from an employee point of view, I think there was a study that had recently been done that an average employee spends two and a half hours a week searching for information, which is astounding. Yeah. And in the average company, over 1000, people have over 137, different sass apps. If you start to boil that down, just if you look at HR and benefits, that’s pretty real and material, right? Because oftentimes, you know, there’s these great systems of record like workday, you know, there’s great systems of record like ADP, right? or other types of systems like crona, UKG, or whatever it might be. And on top of it, they might have a knowledge base. And then you start to compound that, and you start to get to that place of why people are spending so much time with it, because they’re not often sure where they go. And, you know, so what we kind of look towards doing, and what we’ve seen in terms of impact and results is, you know, we look to try to avoid the call before it even happens, but also go to, to the employee in a way that they liked to interface with technology. Right? So it’s really difficult for companies, I mean, the pandemic has been a perfect example, if you think about it, you know, of how quickly information is changing just as quickly as it’s produced. Right? If you think about it, what’s the policy? You know, are we going to go back to the office, right? You know, the whole notion of being able to go down to the proverbial door down the hall and say, Hey, you know, can you help me with this, it’s kind of gone, right? And so people are trying to figure out ways to be really connected, right? And you want a solution that that really reflects a company culture and tone that everybody works so hard to provide for people, but in a way that they consume it easily. And, you know, even if you think about, you know, what we we’ve seen from some of our customers so for instance, we have a very large manufacturing company that has about 50,000 people, and they’re in 26 different countries, and they have 20 different languages. And what we did for them is we supported them A big workday customer there UKG. They have service now for HR and it and a myriad of other systems. And while they started the journey thinking, Okay, well maybe I could look towards using these conversational bots, for each of these point solutions, it wasn’t really going to solve their problem, right? Because it’s that continuation of where do I go for what? Right? There’s the service. Now, I T bot, there’s the ServiceNow HR bot, there’s the workday bot, but nothing pulls it all together. So because people don’t think linear, they don’t say, Okay, I’m going to ask all my payroll questions, and I’m going to say, let me do all my HR stuff. And then now let me take an idea about it. So this notion of pulling it all together, they placed it on all these different solutions that they offer their employees. And really, that idea was what? And so what they saw in terms of impact and results, within the first three months, they saw 60% reduction in terms of what they had in terms of a call center. Right. So what that really allowed was two things, right? It allowed the people who were supporting employees with those mission critical or strategic things that they really need support with, to really focus on those challenges. It’s not about eliminating, you know, the headcount for shared services. It’s about making them more effective. And then the second piece is also that experience, right? So, you know, now I can just really quickly say, okay, instead of me trying to figure out even things as simple as updating your personal information, where you change your address, or for a manager to get a an approval on a time off request, on their on a text, right, make things simple, because that pushes and pulls into those systems, but makes things super convenient. And we’re seeing pretty up to 80%. impact in terms of a reduction of those repeat calls within three to six months, which is meaningful.
Brent Skinner 12:02
Yeah, that is, I mean, he mentioned shared services. And not, the idea isn’t to reduce headcount. But to to make them more effective, right? You have this labor expenditure, you want to be getting the greatest sort of, you want to get the greatest potential value from that, right. And, and so there’s that you’re making shared services more, you know, potent, more formidable, just a greater asset to the organization, but you’re also improving the day to day work life of those Shared Services people actually write because they’re not, they’re no longer dealing with just inundation. And I’ve looked at enough of these, to know that some sometimes these Shared Services folks are inundated with literally inundated with just tedious requests. I mean, it’s nobody’s fault, except the systems, right, the but you know, your employees asking, you know, basic questions, and they’re having to sort of, you know, respond to those all day long. And so, and so you’d probably improve retention in your, in your shared services, staff, and all sorts of things. Can we just get into the AI part of it? And you may have mentioned something earlier in the and I’m sorry, if I missed it, but yeah, let’s get into this, this dive a little bit deeper into what, how the AI works with us?
Melissa Swisher 13:29
Sure. So I, so one of the things that we did, when we started the company is, you know, we, we had some real notions of what our strategic principles would be. And the first was, is that what we wanted to provide companies with a platform versus a toolkit, and I’ll tell you a little bit about that, meaning that we take on all the heavy lifting for you. So, you know, there’s a lot of solutions, where they’ll give you a framework of, Okay, here’s some of these intense, you know, intense meaning that here’s some of these questions. And then here’s these responses that come with it. And they ask to then updated to some of those different variations, and it’s a pre found template. Now, the reality is, is that ways people integrate or interface with these types of tools is, varies greatly, even from one person to the next. And so, with Socrates, we have take on all of that auto processing, that intent mapping is automatically done by us and its process and continues to get better through supervised learning and the power of our customers, right. So, continuing to and we take on all the heavy lifting, from that point of view. So oftentimes, that can be a very resource. Heavy support needs, you know, oftentimes coming for instance, if we’re talking about HR that falls outside of HR to get those resources from for instance, it. second piece that was really about, we look towards supporting the IT strategy without creating any sort of limitations, meaning that Socrates in terms of the platform itself was built in two parts, right? There’s the actual processing engine. And then there’s the front face interface, which is really what you call the bot. But the processing engine, and we’re all the IP is, is that it was built in a way that Socrates doesn’t have to be the front facing boss. why that’s important is because we actually can assign other intents to other bots, but not compromise the end user experience. So it can do bots about capability, which means that from an end user point of view, they’re not going to know the difference of who they’re talking to, it just creates that aggregate experience itself. And the integration engine itself in terms of the ability to integrate to all of these different systems and solutions, we have 90 plus percent of the API connections done for HR, the HR and benefits and IT systems. And then anytime we run into one that we don’t do, it’s always in our best interest to do so. So those are some of the pieces of really what we do in terms of aggregating across all those systems. So it goes beyond just the workflow. It goes across those systems to and can auto detect whatever language they’re talking, right, we support over 108 different languages. Right. So from that global point of view, making it super, super simple to have that experience across the board. 108 languages.
Brent Skinner 16:50
That’s just about all of them, isn’t it? I’m not sure exactly how many there are?
Melissa Swisher 16:55
Yes, I think it’s a it’s a pretty extensive list. Yeah, for sure. Yeah. Yeah. Um,
Brent Skinner 17:02
by the way, I should explain to the audience, yes, I am drinking out of a mug that has the Santa Claus on it, even though it’s August, I just realized that saw it on the screen and figured it just may a couple folks. Anyway, um, Woody, let’s, let’s talk about that. That the, the future, you know, maybe the near future, for this kind of capability in general in the industry? And maybe the not so near future? Let me share with you a couple of ideas that I have. Sure. With AI, being infused into this, right, like real a real AI, not just machine learning, although that’s part of it. But you know, but with natural language processing, plus scheming natural language processing, and these sorts of things. Do you see these systems? Well, I see these systems, potentially anticipating employees questions instantly? If, if Socrates already does this, in a sense, at least, so first of all, this Socrates does do that already? Or are you close to doing that? anticipating employees questions or, or being there and, and maybe from some sort of an algorithm, understanding what the employee is, is apt to ask, so that it takes less time to respond to that? Because you because the system already knew, or had a good idea of what the person was going to ask, but then moving beyond that, so you have let me let me explain where I’m going with a C of a, a, an employee, that’s, and this is all HCM centric, because that’s where I live in play. But let’s say you have an employee, that’s a potential flight risk, right? They may well be looking to leave the organization and they happen to be a high value employee, maybe one of your warehouses will just be you just have no clue. You haven’t a clue how that warehouse will survive, or go on with this particular person leaving, right. And so do you see these systems one day, maybe being able to anticipate the types of concerns and questions such in play would have and maybe leading them to certain types of information that might present this might even sound a little Orwellian, but and persuade this person to stay? right if you know from other aspects of the system that this person is a flight risk? Is this is this really to science fiction or is that a possibility? In the
Melissa Swisher 19:55
great, no, it’s a great question and you know, Just in terms of the experience with the individual, you know, it’s so critical for employers today. So I think that So to answer your question, Socrates, in many senses, is there in a number of those ways that you mentioned? So anticipating what the employee is going to ask? So, you know, in a very basic form, right, if they’re saying, Okay, how much time do I have? And it gives them that personalized? Okay, here’s how much time you have. And by the way, do you want to take any time now? So starting to think about, can we be proactive to what an employee is looking to do? Or also from a sense of, Okay, let’s say somebody just had a baby, their first their first time parent? And how do we look at that, in terms of that, that experience for that first time parent, and there’s so many different promotions, or I’m sorry, programs and wellness initiatives, you guys might, you know, the company might offer? Right? So the ability to cross promote, and say, Hey, did you guys know, congratulations, we’re thrilled for you can automatically ship out a gift on behalf of the company to that individual? It can also then it also can then it can, it can then also push other programs, right? That they would have Did you know, we had a healthy mom healthy baby program, right. So those are all those different things that that we can do in order to really start to promote that. But that next layer, right, of what people are looking to do is really around that personal relationship with the company and that layer of trust. And so, you know, what’s going to warm people’s hearts, right? What’s going to be there for them in terms of the experience that they have, right to make it that, you know, the company cares about me as an individual. And part of what we work really hard with our customers on is to create these cultural experiences, right, whether that’s giving them information that lives as a multimedia type of experience, right? So you know, instead of giving them just a straightforward pop, you know, policy answer really warm their hearts in the same place with either some sort of, you know, meme or video content, in terms of how somebody might interface and also making sure that from the point you mentioned, how do you know if somebody is a flight risk, right from that capacity, also being able to say, also, to think about it from a point of view of saying, okay, here’s the things that, you know, you get records of every question and every single response given, it’ll start to indicate trends, right, to start to see, you know, are there patterns of things that would indicate that this person is a flight risk, right, and then giving that opportunity to have a fairly strategic conversation with them? to lean in to what they’re looking to do? if that helps.
Brent Skinner 23:00
Yeah, so, and, and you, you describe this from a slightly different angle that hadn’t occurred to me, but it’s absolutely applicable. Right? This type of assistant can help inform a predictive analytics set up in an HCM platform that’s designed itself to help predict flight risk. So if you have even more information, you can help an existing flight risk detecting functionality be even more accurate, because you’re capturing information on that employee that’s regarding his or her activity outside of the HCM domain of the enterprise, which is, which is very, very key. Oh, that’s interesting stuff. So she know what’s interesting is the idea that the AI can really help to, to fine tune these sorts of things. And can you describe to me maybe a situation where Let’s be devil’s advocate here or something, you know, is there any sort of a situation where, you know what, you as an organization, you’re not really ready for this type of other solution. This is not for you yet. What can you maybe describe that sort of a constant consultation that you might give to somebody?
Melissa Swisher 24:45
Yeah, absolutely. And the way we look at this is it’s totally a journey, right? So everybody’s usually at a different place. And a lot of times where we start with customers is to try to say, okay, what’s this long term vision? What do we want to do short term A lot of times what people will talk a lot about right, especially is the quality of their content or the state of policies and the information that they’re telling people, right? Meaning that things are all over the place, there’s information that might be out there. That will, that that will, for instance, so you know, you have all policy, a lot of times these systems grow by ad, but they don’t necessarily take things out. So one of the first things we do when we start working with a customer, we do it in two parts, right, we can do it as a standalone opportunity for customers, or we can do it as it’s a standard part of our every implementation. And the first thing that we actually do is we do something that we call content transformation. And we will actually crawl all of the places where policy benefits information lives within the company, as long as the company gives us access to it right. And then we will actually create an inventory, we’ll create an inventory based on category or topic, global policy, subcategory, all those different things. And then we’ll give them a sense of what that looks like. From there, we actually have an internal tool that we call a compliance bot, that we run about 10s of 1000s of questions of the most common questions against that customers content. And the reason we do that is it identifies out a few things that identify anyplace conflict. So if there’s answers that overlap, right, there’s oftentimes what happens is, is you’ll see different policies that are out there that give different answers, right? Or there’s things out there that have old formatting logos, there’s places where there aren’t answers, especially more now than ever, in terms of what people are asking. And then we create that output for the customer that we also then measure usually against their call center data. And because you want to see what kind of content or answers are we providing people versus what people are actually calling about. So we produce this whole inventory for these findings for the customer. And we also then do the content cleanup. So meaning that we make all these recommendations of here’s the things we’ve seen, we can then also then look at things from a tone, right, because if you’ve ever seen policy, it’s very legal ease in terms of the way it talks to people versus you and I having this conversation. And so we go ahead and do all that really give somebody a great foundation, to really start to build on that that corporate culture and get people what they’re looking for. So we you know, and we kind of take that journey with the customer, right, in terms of what what’s most important and what’s most valuable, we definitely see a grow in terms of really looking to give people those personalized experiences. And then oftentimes, what you’ll start to see is in the questions people are asking, they’ll start to say, Okay, well, my printers at a bank. So then or other things that live outside of what we might be supporting from an HR and benefits point of view. And then that gives the customer an opportunity to start to bring in other places of things that employees are asking about. So that holistic experience and making that front and center for where, you know, whatever the employee needs, whenever they need it.
Brent Skinner 28:38
It’s interesting. what you just described me makes me think of them many, many lives ago. I won’t say how many, long time ago before I even knew what HCM was. So this was give you an idea. I worked at a civil engineering firm in the marketing department and we helped them develop proposals for in response to RFPs municipal RFPs. And we had this huge system, I think it was called the mirrors, mirror system and I RS and it was I don’t recall what it stood for the acronym but it was a immense depository of a repository, excuse me of project descriptions and various other things scopes of work things that had been written prepared by the you know, by the source engineer, himself or herself or, or by somebody else in the in the organization if this all there. And from that standpoint, it was very helpful to have that huge archive of already developed stuff that you can rework for new proposals and, and whatnot, right. But at the same time, it was really tough to search through it was very, very difficult and it’s sort of in a perfunctory, on, you know, on, you know, a perfunctory part of our job was that we would be spending a lot of time searching the mirror system for these, these write ups this content. And I just think about, you know, 20, okay, 2020 years ago, 20 years later, right or about, you know that, I think it’s just night and day, you would probably and I even wonder if maybe this company has something like your solution, or some sort of similar idea applied to that system so that so that their, their employees can be far more productive, assuming more efficient, that That, to me is just, it just made me think of the mirror system to be honest with you, they, as you describe, that immediately thought of my days using the dreaded mirror system. Interesting. Yeah, yeah. Oh, I love this, this is good stuff, to be able to, to have it. And again, like I said, at the outset of this, you know, it’s to say that Socrates AI is essentially a chatbot, that that’s very, very over simplification of what you folks do. You’ve done a great job of describing it and, and to have that sort of a system to be able to, marshal just the, the information from several disparate systems and in give it sort of a central reprises repository sort of feel to it. And to give that, you know, give the tools to, improve the, the, the, the readiness of the, of the, the, the available. information, so my computer’s just blowing up in front of me here with alerts. Yeah, to have that there is is just, I think, a huge productivity gain potential productivity play for any organization that that elects to, to deploy such a thing. Yeah. Is there anything that the we missed anything that that, that you’d like to point out? I mean, I think we’ve covered a lot of ground here. It’s been a fantastic, really, really interesting conversation.
Melissa Swisher 32:26
Yeah, no, I likewise, it’s really appreciate the time. So you know, you know, love to talk more, and, you know, but yeah, really appreciate the time today, we’ll be at HR tech for anybody who’s going to be there, but now really appreciate this opportunity. And we’re very passionate about creating these memorable experiences for employees. And so, you know, it’s very top of mind. So really appreciate the opportunity, Brent, thank you.
Brent Skinner 32:56
Oh, absolutely. It was it was it was my pleasure. This has been fantastic conversation session. Thank you so much. Have a wonderful weekend, and really looking forward to making this particular podcast available for our viewers. Great. Thanks, you too.