HRExaminer Radio Executive Conversations Badge Podcast Logo

HRx Radio – Executive Conversations: On Friday mornings, John Sumser interviews key executives from around the industry. The conversation covers what makes the executive tick and what makes their company great.

HRx Radio – Executive Conversations

Guest: Mike Gioja SVP of IT and Product Development, Paychex
Episode: 383
Air Date: November 6, 2020

 

Transcript

 

Important: Our transcripts at HRExaminer are AI-powered (and fairly accurate) but there are still instances where the robots get confused and make errors. Please expect some inaccuracies as you read through the text of this conversation. Thank you for your understanding.

Full Transcript with timecode

 
John Sumser 0:13 Good morning and welcome to HR Examiner’s Executive Conversations. I’m your host, John Sumser and today we’re gonna be talking with Mike Gioja, who’s back again. Mike is the Senior Vice President of IT and product development at Paychex. He leads the development of all Paychex applications, including the company’s full scale human capital management offering. Mike, how are you?

 
Mike Gioja 0:36 Good, how are you, John? Thanks for having me on today.

 
John Sumser 0:39 Yeah, it’s a bright sunshiny day here in California. How about there in Rochester?

 
Mike Gioja 0:44 Same thing, very unusual. 70’s and full sun. So we’re enjoying the next few days of this weather for sure going into the weekend.

 
John Sumser 0:51 That’s great. So take a moment and introduce yourself in some detail. Who are you, how’d you get into the job? And what are you doing these days?

 
Mike Gioja 1:01 Sure. Hi, everyone. I’m Mike Gioja. And as John had said, I’m the Senior VP here and CIO at Paychex. I’m responsible here for all corporate applications, all IT projects, the security and the seaso office as well. Obviously, we focus a lot on all of the key technologies in the space today, and we’ll talk about some of those. I’ve been in technology for over 40 years, and I’ve been in HCM for over 20 years. Prior to Paychex, I’ve held some key positions in IBM, AMEX, and then Fidelity SAP, and PeopleSoft all focusing in HCM space.

 
John Sumser 1:35 So you are a tried and true veteran, just in case there’s somebody online who doesn’t know about Paychex, what’s paychecks do, and how’s it different from the competition?

 
Mike Gioja 1:45 Sure, Paychex leads the way by you know, taking complex HR payroll and benefits and making them simple through our innovative platforms and technologies and our highly trained HR consultants around the country.

We have our all in one HR payroll Time and Attendance benefits platform, we call that Paychex Flex, which helps our clients obviously support their employees in the office or working remotely, we could be talking about that today, I’m sure by managing critical communications, keeping them all engaged and helping them stay productive. We’ve got millions of monthly users, we’ve got, you know, over 30, plus self service transactions for employees, and obviously a lot of capability for the admins. Paychex also helps our clients we have recruiting and talent management capabilities as well. We also have retirement and section 125 plans. And we’re also a major player in health insurance being one of the largest brokers in the US as well to help our clients protect but they worked hard to build. And Paychex is really has a 50 year history. We have over 680,000 clients and a tremendous amount of data and insight to help underpin a lot of decisions, the products and things that we do today. And I’d say the last thing we’ve done is over the last decade or so we’ve been really undergoing complete digital transformation. This is everything from our SDLC. And how we do everything in our SDLC software development lifecycle, obviously working with advanced architectural concepts in the platforms. So we do all of that today, obviously Service Automation, AI, ML and chatbots. And all those things we’ll talk about today, and RPA. And of course digitalizing our sales and marketing processes as well.

 
John Sumser 3:24 So what don’t you do this? That’s quite a comprehensive portfolio? What don’t you do?

 
Mike Gioja 3:30 Yeah, we pretty much touch everything you can think about in HR payroll benefits from hire to retire.

 
John Sumser 3:36 That’s pretty awesome. So I imagine that COVID has has dramatic impacts of both the ways that the company operates and the way your customers operates. Tell me that story.

 
Mike Gioja 3:47 Yeah, it’s been a heck of a story for everyone. And many, many companies of all different sizes, of course, struggling with these issues. So you know, obviously, the first thing is when we got into the COVID situation is ensuring the safety security of our own employees and making sure that that was our top priority throughout the pandemic and is still our top priority, obviously.

So back in March, we implemented our you know, business continuity plans. And that meant that we had to move 95% of our employees, and we have about 15,000 employees in the US and Europe and we had to get them to work from home within five days. That was the key goal that we had to hit, which we did. That meant by the way also we shut down about 100 plus facilities around the country, and only for essential personnel that needed to do a critical on site. And that’s probably today three 4% sometimes 5% at most. Also the HR team has been working diligently every day. There’s a you know, major cross functional team of HR marketing and some product folks to provide tools, resources and communications and all the things necessary to support the mental, physical and emotional well being of our own employees. Obviously as being very key and we’re constantly you know, keeping in touch with them as they’re working from home and dealing with all the stresses all the families are dealing with with kids at home and everything else that you can imagine in this environment.

We also have to ensure the availability or even resiliency of a lot of internal technology that we have. And it’s very good. Unfortunately for Paychex, that obviously we’ve been focused on those technologies internally. For example, you know, virtual private networks, WebEx, we utilize Cisco services, we have WebEx collaboration tools that were already implemented, that we’ve used, essentially, for people to collaborate across all the different buildings and different sites around the country. And so we had to quickly move them at home. And what we did was we were able to just get 7000 soft phones, for all the service folks that use our unified communications, and headsets, sent them to their homes along with their desktops or laptops that they needed. And they are productively working from home using all of the same tools as they would be in the office. And we can also see how they’re working and do all of the normal metrics and everything that we would capture that a service organization does, because they’re connected into the soft phones and right into our network properly authenticated, and secure. That was another thing you have to really look at in the work from home given sensitive data, all the security aspects about it, and managing all those things as well.

So we successfully got everybody working from home, and we keep in touch with them. HR continuously provides tools, we do little surveys and checking in with everyone and doing all the leadership obviously checking in all of the time to see how people are doing as we’re still progressing through this pandemic. In relation to our clients, obviously, we needed to quickly respond to many challenges. Obviously, the biggest thing that was coming up then was the PPP reports, we put out what we think was highly differentiated capabilities and very simplified forms, integrating automatically all of our content, and payroll and HR information that we could capture, along with giving the clients the ability to enter all the things related to the PPP with the rent and mortgage and those things. And they were able to take those forms to the bank. And the net of it is we expect that we’ve seen about $28 billion in PPP loans go to the Paychex client base. And then we put out a very easy forgiveness calculator and signature ready forms. Obviously, behind all this luckily, through our SDLC in our continuous development and integration process, we were able to quickly and immediately create these and deploy them and enhance our infrastructure. Because as you can imagine, like these things with the PPP, as soon as those things opened, we knew every single client would be logging in at the same time, and you need to make sure the infrastructure was there. So you know, we just were able to quickly scale up based on our thoughts and create more containers and get everything out.

Now one other set of things along with that is that a product team came back to us and asked for a number of enhancements to help our clients which we quickly were able to deliver. Some of those are HR conversation so that the company, the administrators and managers could do pulse check ins, with their employees through HR conversations and be able to track and record that. HR Connect, which allows the employees and managers to submit cases, workplace incidents COVID incidences and then seamlessly transition those to OSHA reportable instances, as needed. I’d say the area of document management, we did a bunch of work for automate employee distribution of policies and procedures. So we had been working on in just delivered document management repositories for our clients and AEC capability, we gave them the ability to create policies, documents, distribute them automatically, and then track the signatures and the consumption of that material. And then time management capabilities. And then obviously, one of the key ones too, was going to real time payments, so that we have the ability to quickly process a payroll at any time and immediately fund it to the employees counts within seconds. So those are some of the major drivers.

 
John Sumser 8:48 So what were some of the things that surprised you in the process? What was hard? What were the challenges that you encountered and what was hard that you didn’t expect to be hard?

 
Mike Gioja 8:59 I’d say, you know, everybody ran into logistics issues, because when it hit a lot of issues, so when I look at it internally, it’s like, okay, where can you grab headsets? How do you get the soft phones? What did we have? As we started to see the pandemic issues start, my team was great. They went out and said, we’ve had to lineup some things here, they were surprised at what they were hearing as far as delivery times. And we were able to work through that. But that was surprising and a struggle. Because all of a sudden everybody was trying to get the same stuff. So logistics and getting supplies definitely was something that we didn’t expect to be of that nature.

And then, you know, we know with regulations, there’s always churn and change with the regulations. But this was beyond anything. As the company the country was really trying to solidify stimulus packages and changes and you get a lot of indications and implications, not a lot of specifications. And I would say probably like on PvP alone, we probably had 60, 70, 80 releases in a short period of time. That’s how quick we had to move by constantly tweaking and tuning it with every clarification we got, but at the same time giving our clients something that they could see and understand because they’re dying for information. And then marketing on product got together. And really, I think it was amazing to us how much work it takes for that cross functional team, which I think has done an amazing job in understanding what’s going on around the country, how to put that down in a way where our clients can consume it and come to Paychex.com and understand it, let alone then how do we take that and figure out what else we can do not just from a service point of view, but an information point of view. We put a lot of content into our chatbot, because we could see a lot of questions coming in. So that worked at the cross functional team did in understanding the landscape and the regulations and what’s happening around the country to pull that together, put the knowledge into the bot, put it onto Paychex.com. I think that was more work than anybody expected. And with a lot of full day and nights with the constant change and learning that everybody’s was going through. It’s just amazing. That’s kind of settled down. But I expect it’ll pop back up when you know, they hopefully readdress stimulus pretty soon.

 
John Sumser 11:07 So that’s a great segue, because the next question is about how you’re thinking about AI. Let’s start with chatbot. Chatbot means all sorts of things to all sorts of people. So when you say chatbot, what are you talking about?

 
Mike Gioja 11:20 Well, okay, our chatbot, we call it our flex assistant intelligence engine is the branding we’ve given to a whole set of a tech, let’s say intelligent technology that works together, which leverage NLP obviously and ML. So are flex assistant is the key component. That’s the chatbot. It’s integrated with our Help Center, which provides enterprise search across all of our content across all of our ancillary so the chat bot appears, wherever you are online or on the phone or mobile, wherever it appears, then you can engage that chatbot, and it understands where you are, and then asks you an initial set of questions based on that context. Also, we use Pendo. We obviously have robotics for RPA, we have voice assistants, we have OCR. So those are all the things inside the intelligent engine, the bot itself today, the majority of the chat interactions with the bot are 55% end up being resolved by the bot only, meaning we can also take you to a chat, a live agent chat and give you that ability. And then you can actually go from the live agent chat to a call. And we have complete traceability and all of the content, and everything that goes along with that. So if you end up going that entire stream, the person you talked to on the phone has seen a look at all of the interactions. And I’d say that’s the key separations that we have. And then the bot obviously is really more of let’s say, a question and an answer approach. So it’s consistent regardless who’s asking the question, we take the question, we try to understand your intent, obviously, and then we’re giving you an answer. And we allow you to do multi turn in that process utilizing the Google NLP framework.

 
John Sumser 13:01 That’s great. It sounds to me like this would cause an upheaval in the way you think about doing customer service and staffing the various levels. So you talked about the chatbot, being able to solve 55%. And then the other 45% gets passed through some kind of escalation, starting with chat during the calls. How do you staff that? And how do you predict workloads on that. This is a bugaboo, and a lot of chatbots and my sense is that you’re doing it well. So how do you take about predicting the volume of work in the call center, when you start automating it in the way that you’re doing?

 
Mike Gioja 13:40 That’s a good question. Obviously, what we do, because we’re always concerned about customer service and responsiveness as we start conservative. So as we do things with the bot, and we work with service on content. So let’s take for example, 401k, or payroll picking area, we sit down with the service providers that handle the majority of the questions for the clients, we ask them what they think are the questions that they wish they didn’t have to deal with, there was a fellow straightforward, where do they want to do consulting and advising. And that’s how we start with putting content into the bot. And then you know, we through Pando and others real lead clients there once in a while to let them know that the bot is there. And that’s something that they could use. And then we watch and we model. So as I mentioned earlier, we’re on this unified communications omni channel framework through Cisco that everybody uses now at home successfully. With that, we get all of the interactions coming in through any channel, right? And so we understand and get a bunch of metrics and understand all of the interactions that are being done by client through all channels. And we look at that and we watch those trends compared to what we do with the bot and we see bot activity. So as we see clients and employees knowing that oh, I can ask it my 401k question. Oh, it gave me an answer or now actually, it’ll be like release You write to the Self Service transaction to make the change through deep links. And as we see that, we can start to see the bot interactions going up. And we can also see if any Unified Communication interactions are actually going down. Or we are definitely seeing that in major subject areas. And we’re seeing that trends continue on that path. And we expect there’s a point where it kind of flattens out. And we do this by all subject areas, and then we determine where we need to go. And we also look at it, let’s say very cyclically, because we now know and we’ll put in new content for year end, because admins and payroll folks will be thinking about year end and changes in the new year and things so or open enrollment, which we just went through the most of the base. So based on where we are in the cycle will also promote and add additional content for people to understand any changes in that space as well. So we try to correlate a lot of data, we look at that. And we take Google Analytics, Unified Communication analytics, chat analytics, and then all of the stuff to see how the intents are performing. There’s a team constantly looking at that on how the engine is responding, and we tweak accordingly along the way and watch the trends and adapt.

 
John Sumser 16:10 So that’s an enormous undertaking and Paychex is a big company, it makes me wonder whether or not you think that littler companies will ever be able to handle this sort of level of automation that you guys are getting. What do you think?

 
Mike Gioja 16:26 Yeah, if you’re a small company that’s harder to do, the technologies are evolving pretty quickly. So I do believe AI, ML, or chatbots, that takes quite a bit. But you know, they could also benefit from RPA. We’ve done a lot of back office robotics work today, we’ve now got 16 categories where bots are doing work. And it’s already saved over 150,000 hours of work for our service providers. So there are elements that maybe small businesses can do and get into. And there are growing technologies that we’re constantly looking at, in AI and ML, that they can be expensive, some of the tools and hopefully that will change over time. But they’re you know, the yard developer friendly and not too difficult to kind of pick up and start to go with, it’s just a matter of how many different things you may be having in your back end that you’re going to try to connect. So the simpler your environment, the easier to automate, the more complex obviously, you got a lot of use cases to watch for. That makes sense.

 
John Sumser 17:22 Yeah, makes all the sense in the world, I’d like to get a deeper look at how you guys think about AI. And it’s interesting to start from the idea that what you laid out is the way you figure out what to automate next, and where to move next is by examining the data from your last adventure, from your last trunch of technology. So you’ve got this interesting system that you’re pulling together that tells you where to make the next move. Beginning with that, then what’s the rest of the approach? What drives that engine that you’re building?

 
Mike Gioja 17:56 Well, I think the whole idea behind the engine intelligence is to get to a point where users can pick how they want to interact with the system, how they can learn with the system and facilitate that process as much as possible. Because that was one of the main designs of the integration Help Center because we say, Well, some people like to read, some people want to watch a video, some people want to watch just small snippets and do a particular transaction. Some people like just take me there and let me do it myself. And hey, you know what, I just want to tell you, and can you do it for me? Those are where we’re evolving. And so but the Help Center does is based on how you interact with us. And what you tend to do, we tend to say, Okay, this is your behavior. So as we see that, and you have, let’s say you’ve asked us questions in the last six months, and we see you typically tend to go to Video clips, well, we’ll start defaulting you right there and then give you an option if you want to go somewhere else. So that we can quickly respond and the way that the user needs to be able to do what they can do as much as they can on their own and interact and communicate whoever they need to on their own. So that’s why we constantly continue to evolve the technologies as well. Like recently, we just put out voice assistant. So now you could ask the system to tell me what tasks I have. And it’ll tell you what tasks or employee can say, hey, did I get my paycheck? What was my you know, what’s my paycheck, and we, it’ll tell you what your paycheck is, you don’t need to go online. We want to facilitate all different ways. So the way you work effectively as an individual is what we want to help you be as productive as possible and support those channels,

 
John Sumser 19:29 Cool. I’m getting the sense that your focus is in your existing swim lane we do these things. And COVID has raised a whole bunch of new issues that you’ve dealt with with your own staff. I wonder if you’re extending product functionality out to do other COVID related things, tracking people tracking cases, that sort of thing?

 
Mike Gioja 19:53 Yeah, we did do that. We have definitely done some of that, as I mentioned with our HR Connect, which allows you to have cases and track cases. And submit them to OSHA and use those forms and HR conversations, obviously, where the world is going given HR is getting much more active about the individual well being of their employees in and out of the workplace, paychecks as a company has always been very focused on that. I mean, one of the things that’s very interesting going on right now is there’s a lot of new business starts, huge amount of new business starts going on. And I believe there’ll be a whole bunch of business starts in this area of well being bringing technologies into the workplace, and other ways of which we need to quickly try to understand what those are and begin to enhance more are HR offerings to provide the capabilities to help the HR organization understand and track, let’s say, much more than done anywhere today, right, the individual health and well being of their employees, and potentially detect things that could be problems, which they can then proactively assist with. And I think there’ll be new things coming out to help with that, and lots of innovation to occur. But that’s certainly a change. So we’ve reacted with some things to help immediate needs. And then obviously, from a roadmap and planning product is very much in the depths of of looking at what’s the next set of things to do.

 
John Sumser 21:12 That’s great, so you are dancing on the edges of all sorts of ethical issues in sort of this robust application of AI to the world and the consequences of the world that we’re in. What are the big ethical concerns in your work? And have they changed as the result of the social climate and pandemic?

 
Mike Gioja 21:34 Yeah, there’s a number of things. And I’d say another element I’ll touch on here, too, is privacy, which you see more and more as well, which has a lot to do with the data you capture about people and the you know, California, starting with ccpa, with the California Privacy Act, more and more is going to happen in this space, and you’re gonna have to be more cognizant, so that’d be another point. So I would say ethics is obviously always been a very key part of the paychecks culture. And we’re very proud of winning multiple times, probably the dozen times or so, earned many awards, and always getting placed very highly on the atmospheres Institute of the world’s most ethical company. So we’re very strong on that. It’s clear, what we need to do and have been doing is work hand in hand with the services team and all our service providers in the different lines of business to learn from them. What’s the type of work that they prefer not to handle? Where do they want us to automate. And then as we drive efficiencies, utilizing that, to help with work life balance, and stress management is the primary focus and then company efficiencies as the secondary focus. That’s the way we look at it and go very conservative and very slow. As part of that approach, the bot, as I mentioned earlier, it doesn’t provide guidance and answers questions in the same way for everybody. So that helps us right now keep very clean, we specifically avoid use cases that would drive us to any kind of bias or having the bot begin to train itself on how to better answer the question next time because we still think we need to have a fair amount of human factor in this. So we haven’t designed bots to have personas to understand your tone and change your tone. We’re keeping it very consistent, and building that out.

Now, I’d say a couple of areas that are key concerns and some of the things that the way we look at is that, especially with privacy mentioned and where the world has gone is we want to make sure that the employers protect themselves from their own bias. What do I mean by that, we want to provide insights to them that are intentionally not let the clients tweak the model. So we want to give them insight. And we want them to trust the insight to give enough information for what we’re showing them. But we don’t want to open it up for them to tweak it potentially then with their own bias, even though it’s probably an unconscious bias. And then how do we then give employers an accurate model without creating liability for them, we want to provide access controls obviously, is key that’s always there. But we also are thinking about how much data do we disclose about what are all of the elements that are within our model? We want them to know that there’s enough data there to trust the model, but not necessarily give them all the details. So for example, maybe you don’t want to disclose that gender is an attribute because then maybe they might think differently or want to do something or better get at that.

So that’s another thing that we’re constantly assessing is not only how we give something that we feel accurate with but how much control are not that we give to our client and how do we make sure if they don’t end up doing something unconsciously putting themselves at risk. And we always have to balance transparency against the risk and trust is the foundation at the end of the day. So between that model and how we interact with them from a human point of view through the user experience, hopefully will drive those things out. And I’d say that last point if I want to come back to the ccpa what this means going forward for those that aren’t familiar with The ccpa is in 2023. And this guy ratified and this election for 2023, I believe, is essentially that employee ease of a client that we service or other people might service will have the right to ask what informations you have about me. And I don’t want you to keep any of this personal identified information. And I don’t want you to promote or do any of this specially, so they could have more control of their information. And we have to have the ability to pull that information, especially if that client leaves us or that employee leaves the client and they want everything deleted, we have to be able to do that. So now you have to look at and go, what does that mean to AI and ml. And we have to make sure through our data pipeline as that data goes through. And those data models that are out there, and the data sets, data, sinks, data breaks, all of those things that are out there, we got to make sure that there is no way that that content than they are can be tied to any identifiable individual. So we need to abstract the information, not bring over the personal identifiable information, but make sure we’re capturing the right trends and attributes, so that we could still maintain the models and maintain the privacy of the individual. So that’s another thing that we’re looking at as well.

 
John Sumser 26:16 Awesome. Awesome. So we’ve blown through our half hour. Is there anything that we didn’t cover that we should have?

 
Mike Gioja 26:23 Oh, no, I think we hit a lot of good interesting points, John. So no.

 
John Sumser 26:26 It’s an interesting story, you’re approaching the problems in a differentiated way. It’s awesome. Would you take a moment to reintroduce yourself and tell people how they might get a hold of you?

 
Mike Gioja 26:37 Sure. And thanks for hosting me on your podcast today, John. Hi, I’m  
Mike Gioja Senior VP at Paychex and please, if you’d like to get in touch with me just send an email to Mike at Paychex.com or can always connect with me on LinkedIn. Thanks very much for listening today.

 
John Sumser 26:53 Thanks again, Mike. Thanks for taking the time to do this. And thanks everybody for listening in. This has been HR Examiner’s Executive Conversations and we’ve been talking with  
Mike Gioja of Paychex. See you back here same time next week. Bye, Bye now.



 
Read previous post:
2020-11-09 HR Examiner article Assessing Values in Online Technology Ten Things We Learned stock photo img cc0 via pexels roman koval 3300448 sq 200px-p1.jpg
Assessing Values in Online Technology Part 2

“In the world of psychology, there is a view that a correlation exists between personality traits and performance (behavior). After...

Close