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: Jamie Troiano, founder and CEO, PredictiveHR
Episode: 329
Air Date: June 28, 2019




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

00:00:13:16 – 00:00:28:14
Good morning and welcome HRExaminer’s Executive Conversations, I’m your host John Sumser and today we’re going to be talking with Jamie Troiano who is the CEO and founder of a company called PredictiveHR. Jamie, how are you this morning?

00:00:31:14 – 00:00:32:09
I am doing great, thank you for having me John.

00:00:32:26 – 00:00:40:15
Yeah. Yes why don’t you take a moment and introduce yourself and tell me about your background, how did you get here? And, well we’ll move from there.

00:00:42:05 – 00:01:12:25
Sure. So my name is Jamie Troiano. We’re based here in the Boston area and I have a background in finance and then I moved into HR. I’ve been in HR Operations for close over 15 years. And one of the things that I always was focusing on is. Recording and analytics and getting disparate HR systems to talk to each other and

00:01:13:10 – 00:01:44:00
So every place that I went to throughout my career I always ran into that as being teams. Same problems I would have large teams offshore putting together reports, stagnant reports, because they’re a snapshot in time, trying to normalize the data and then bringing giving me back reports that executives were making business decisions on, critical business decisions, on data that fail 3 4 weeks old. And because it was a snapshot in time.

00:01:44:00 – 00:02:04:27
It was some of it was wrong. So for me I went to a number of different companies. I tried to get someone to build me what I wanted. And no one would do exactly how I wanted it. So I ended up bringing in a very close friend of mine Charles Occhino who is our CTO. And we’ve we built the product together.

00:02:04:28 – 00:02:15:10
Tell me how you fell into HR. Nobody wakes up in the sort of sandbox when they’re five years old and says Mommy mommy I want to be an HR guy.

00:02:17:10 – 00:03:12:14
Yeah it’s actually a very very funny story. But like I said I started I have a finance degree I was working for Thompson Financial on their financial software so I I actually learned technology became a certified SQL server DBA and then I way I actually went back to school was getting my master’s and I took some time off that wanted to go full time and just get it done and my wife became pregnant and she’s like you got to go back to work and a friend that I played softball with started up a recruitment agency and he allowed me to work part time and while I was finishing up school and she had a good approach where because I was the hiring manager he wanted I could talk I could speak differently to both customers and candidate because I was in there living it doing technical recruiting.

00:03:12:15 – 00:03:17:18
So that’s actually I wrote in was in the recruiting side huh.

00:03:17:21 – 00:03:42:23
That’s a it’s an interesting place from which to approach a. It’s a different world where because it’s the market facing part of the company rathole and markets it behaves really it was funded deliberately as a resource so it’s always interesting to hear what it was like to go from being in recruiting to to being in the risk the.

00:03:43:12 – 00:04:20:10
Yes. So what was funny. So I started out in the agency and it was a good way to get my feet wet and understand that you know they taught me how to recruit. But what what happened was even there. They’re like oh you’re technical here take our eight yes and fix that because nobody likes theory yet nobody their. They’re never done right. So everywhere I went. So I started out doing contract recruiting when I left the agency world and everywhere I went they’re like oh you’re technical just take our ATF. So I quickly got into operation started in DEA operations and then moved into the bigger ITAR realm.

00:04:20:12 – 00:04:24:24
So it’s funny how it evolved over over the year.

00:04:24:24 – 00:04:42:25
Yeah it wasn’t great with a great time because I would guess I would guess that the idea of future operations didn’t exist when you started working the revolution to get to the idea that nature should have an

00:04:44:14 – 00:04:44:26

00:04:44:26 – 00:05:16:19
It really did. I mean there was there was usually in it I think back then really I teeth alone. So really was an even like an H.R. I asked manager. So it was very easy having someone with my skill set being able to just jump in and just take control of the system. Especially at that time and even say the first thing that you go to a recruiter or an H.R. person the first thing as they’d always say is like I don’t like my Yeah I don’t like my system it doesn’t do what I need.

00:05:16:22 – 00:05:33:24
And honestly when it comes down to it the systems usually can do what they want. They’re typically just not configured correctly. So that’s really where I started getting into it and reach configuring systems to make sure that you’re fitting their business process.

00:05:33:25 – 00:05:53:16
That’s really interesting. So why do you think systems are configured properly so they’re installed by the vendor. There’s some huge frustration that we’re going to get right at the beginning and end yet. What you seem to be saying is that the universal problem with atheists is that they’re installed.

00:05:54:13 – 00:06:30:21
Yeah. It’s just not Yes as if any. And part of the reason is typically though. When the vendor comes and don’t talk to you about. Oh we’ll use our consultants and they really understand their system but if you look most of their consultants have never actually been practitioners so they don’t really understand it from the business side. And to me you have to have that business understanding of how people use systems to be able to configure it correctly. So that’s one of probably the biggest issue is they know because they come in and they say yeah we’re gonna do it for cheaper.

00:06:31:06 – 00:07:02:02
We’ll give you our consultants and they know our system inside and out and and they’ll just set it up for you. One is they don’t tell you how much work it really requires from your internal side. I mean you really need a couple of dedicated resource just doing all of that finding all the requirements making sure they’re right. Handing them over make sure they’re configured correctly and testing them out. This is a massive undertaking and most of the time companies because they’re not prepared for that.

00:07:02:12 – 00:07:31:01
Everyone is still on the project are also doing their day job so they don’t dedicate as not high in that they need. So this is really where we started out was for me doing system implementation I before we even did the product we were doing some configuration implementations and this is why it’s so critical that we still have these arms Bay on the consulting side because it’s a very kind of you know sector system.

00:07:31:07 – 00:08:06:20
Lots of interesting notion. So when you go and you get involved with the new company’s system you’ve just described the configuration process that basically requires the vendor to understand the the actual work that gets done in the department in a unique way that looks like some sort of very detailed department level job analysis right. And those sorts of studies are expensive and time consuming which is why they don’t really get done so.

00:08:06:20 – 00:08:38:11
So you’ve got a core expertise. How do you make that. How do you make that process digestible where your client does that it’s still the same level of work and understanding on both sides of the equation. You have some sort of cost advantage because you have core expertise in actual use of the systems but every company different. Right. And so you have to go out to find the differences right.

00:08:38:20 – 00:09:11:18
So what we do when we approach a project we always break. We don’t just bring in one person we usually bring in a team. So we we provide you with experts in specific area. So we always bring in our subject matter expert. Each are subject matter expert to lead the project but we have a benefit expert a payroll expert that we augment so. So really even though you’re paying for like one person. You’re getting like bits and pieces of five different expert.

00:09:11:24 – 00:09:41:28
And so we really get and we really dive into is all of the process your work is the company’s process. We get into the level of detail we’re understanding what’s working what’s not working because when you’re changing systems this is the opportunity that you have to fix the processes that are broken. So if you go through and you fix those identify those processes lay them out then you can go and start configuring the system to meet your process. Yamato.

00:09:41:29 – 00:09:56:04
That’s the way I was going to say that’s the way I have always approached it when I was in the operations side is is I want to find my process and I want interest and then I build my infrastructure to meet a process.

00:09:56:17 – 00:10:21:00
So now you’ve got a company that does this and we’ve possibly we probably could go all the way through this country should that never really talked about the part of predictive nature because right. Right. Because because because you you launched solving a problem that lots of people have that they would love to believe they could have solved in a better way.

00:10:21:06 – 00:10:59:05
But that’s not actually the core of the business is that this is something we started with because when we were starting to launch the product the first couple customers that we had we realized their data was so bad that we could we could we had a hard time buying it together so we ended up doing our first our first couple of customers. We ended up doing implementation first or we left the product work. And so then once once the data is in there and it’s a relatively bad date that’s when our product can take over and and then that’s what I predicted.

00:10:59:19 – 00:11:29:19
Oh that’s that’s very interesting. So. So what you’re saying is or the wizard idea to what you’re saying is that is that in order to actually install intelligence in your H.R. department you first have to go through the cleaning and reorientation process the data governance so that you actually got some data to work with that that’s what you do with the implementation part is address that. That’s not right yeah.

00:11:29:19 – 00:12:05:13
If they need it. This is really only in certain scenarios where their data is really bad. The great news this is part of our A.I. is we can do a lot of cleaning on our side and actually produce data government’s plan. But the problem is that we always have to be fixing it. And sometimes it’s better to just work with the customer and fix that right at the source. That way they understand what they’re doing and they’re they’re making and constantly making improvements than theirs and they’re the way they do.

00:12:06:20 – 00:12:37:10
That’s interesting. So so part of what you’re talking about is there is an already existing but not often funded role inside of H.R. departments for being in charge of data integrity. You might be way of thinking about that scrap. That’s interesting. I think you’re right but I don’t think I’ve heard anybody say that right. That’s the that’s the that’s an interesting thing. So. So now you’ve got the systems are cleaned and perking along.

00:12:37:11 – 00:12:46:21
You’ve solved the chunk to use to those kinds of problems you’re giving a flow of data that you want. What what’s this predictive to actually do.

00:12:46:21 – 00:13:17:04
So this is really where the math is what we do is we can take we’re 100 percent diagnostic so what we do is we connect directly to your system. Whether you have a suite that’s all in one or if you will if you decide to go to the best of breed group and if you do the best in breed. What’s really nice is because we’re pulling the data from all the different systems. We’re actually able to report across the entire ecosystem

00:13:17:22 – 00:13:21:05
And therefore look like you have just one suite of product

00:13:22:10 – 00:13:27:19
Oh that’s interesting. That’s really doing so. So again what does it do.

00:13:27:20 – 00:14:00:27
So what we do is because often what happened in this persistent kind of go back to my experience is typically when you have you have all these disparate systems and they’re working everybody’s working in silos so you’re only looking at one piece of the data. So you’re creating reports and you’re like All right great. Here’s my here’s my head now and then you go to finance finance finance comes up and they have a different headcount and so when you’re young you have your CFO and you’re in your CPO are sitting there talking to the CEO there.

00:14:01:01 – 00:14:30:07
They have different headcount numbers so what we do is by being able to connect the system all the systems together normalizing the data clean it and then be able to present a single source of truth and allows all the executives to be able to speak and use the same numbers when they’re making business so that sounds like the consequence of the work that you do to give implementation.

00:14:30:18 – 00:14:34:27
But the most productive way to show what is for sure.

00:14:34:29 – 00:15:09:12
So for us where we spend the majority of our work right now is predicting around attrition. So we are very unique in this state. We do not most companies that are out there will talk about how they use a eye to make predictions. A majority of them use only trending analysis which is a very loose version of A.I. but for us we actually do really get down into the details and so we we actually go through and use two different flavors of A.I.

00:15:09:12 – 00:15:41:26
. We use the first one we use as one of area where we basically look at every single data point and when we’re running it through our Monte Carlo simulation and it’s the computers determining whether it’s statistically important or it’s or it’s not important. And once we have identified those areas that they are important we run it through a second of a recursive feature elimination RFA where they start combining the elements together and coming up with even a higher statistical value.

00:15:42:06 – 00:16:09:26
And so we really get into the details then and when we pick. So when we predict who’s leaving and you see our forecast numbers those are our attrition is going to spike. I can also give you a report that will tell you who’s at most at risk and you can actually see why we’re picking that they’re at red. So when we add up. So it allows us to be able to take action to help say so.

00:16:10:03 – 00:16:22:14
So as I listen to you describe the process what would occurred to me is that you probably didn’t build the or that you’re using it was probably Google tones like what’s.

00:16:22:15 – 00:16:39:03
So there’s so there’s 10 standard flavors of ice that we that are available. So what we do is we we take versions of that. Those are our starting points but then we lose we use our own coding to adjust that specifically to H.R. specific.

00:16:39:25 – 00:16:56:02
OK. Yeah. But the core real that you know I don’t I don’t know that it’s necessary to go down that particular rabbit hole. So you’ve got you’ve got attrition predictions. What do you do with that. What does a customer do with them.

00:16:56:15 – 00:17:35:02
So so a lot of times what they’re doing is they’re they’re looking at us from to understand. All right Cube who are my employees most at risk and then because we’re looking at this from a holistic standpoint we get it we can say we can identify are these people high performers are they. Are they critical to business continuity. We identify who are those key people that they can’t use. They can’t lose. And it gives them the ways to combat how to keep them with the company because they’re so because when you’re going through it’s not always it’s not always about money.

00:17:35:03 – 00:17:53:07
Sometimes it’s other things. It’s understanding you know that they they want more responsibility they’re taking more management maybe they’re they’re doing one one type of work and they’re really interested in something else and you see in your ELA math that they’re taking all of these other classes because they’re expanding their skill set.

00:17:53:20 – 00:18:14:00
So are you saying that when you do attrition forecasting you provide recommendations from other router systems to go about things that they might do or are you saying that because somebody gets a notification the Jamie is likely to leave and we really need to have the team then the manager could go through those things.

00:18:14:22 – 00:18:43:02
No. So because we’re connecting directly to the systems where the systems are prompting to say hey here’s a person that risk and we have noticed and we have noticed that they’re taking other classes that are outside of their skill set. So this gives the magnet what we’re trying to do is get the manager the ability to to start conversation and have the data to back it up to understand how to take that comfort. Do you see any ethical problems with this

00:18:45:16 – 00:18:51:15
No because we’re just all we’re doing is pulling in data that are already available to them.

00:18:51:15 – 00:19:16:12
If they went in and manually did it we’d just automate prospect let’s assume but I haven’t really thought about this in this life. But I wonder what level of expectation that employees know given that the fact that the data hasn’t been used before and whether or not this interrupts the expectations of I don’t know I haven’t I haven’t thought of it like that.

00:19:17:06 – 00:19:43:02
Yeah I mean I realize you know I was just saying for us that it’s more about a way to connect all of the system together and provide a single view into all of your systems into your and provide that holistic view as opposed to typically companies are making business decisions and very narrow scoped both siloed approach in yeah.

00:19:43:05 – 00:20:16:15
So so I guess what I’m getting at and this is this has been a topic that’s been emerging this summer I think is is the consequences of installing a technology are often not understood very well but because anything easy when you install the changes the way that people work and the changes the way they relate to their organization. And so what you see what you start to imagine is so I just assumed if I took some courses I’d be taking some courses I didn’t realize my boss was going to judge me for doing that and well not.

00:20:16:18 – 00:20:53:23
Now that I know now that I know that I know that I know that I’m being more closely surveilled than I thought. I think that a significant number of people probably start to use that as a signaling mechanism. Right. So once you know how raises are distributed it’s kind of predictable behavior that people will game the system. Right. And and so so I wonder how you think about noticing that five figure out that they’re taking courses outside of my discipline causes my boss to ask me if I want to raise or needed something else.

00:20:53:23 – 00:21:01:06
I think I might start taking courses outside of the system. Right. So and so I’d just ask you have you thought about this.

00:21:01:06 – 00:21:25:11
Not only did you not know so well. The thing really is that’s a long one that you’re only making has very small and there have to be other identifiable attributes that we identify things that people can identify with those other attributes are and manipulate them as a way of signaling economic

00:21:26:06 – 00:21:27:05

00:21:28:21 – 00:21:31:27
Right. And this is how people do right.

00:21:33:13 – 00:21:47:03
And so the answer is the question is when your tool becomes a communication system for some things that are other than it was intended to be used to communicate. How do you handle that

00:21:49:04 – 00:22:00:18
does that always happen and always happens with just the right. It always happens with suspects that they become useful for reasons other than the designers and the honestly I.

00:22:00:20 – 00:22:04:27
This is the first time so I I don’t have any answers to that.

00:22:05:00 – 00:22:19:01
That’s right. I don’t really mean to put you on the spot and I appreciate the fact that you’re willing to say you don’t know because there’s no shortage of people who could writing a book because you went through this.

00:22:20:06 – 00:22:36:18
Yeah you know listen I’m a very straight shooter as I’ve been in this I don’t know what I don’t know. I will tell you but I will go and find out the answer. I just we haven’t thought of it like that because we actually have a very different take than to what you’re proposing.

00:22:37:03 – 00:22:39:21
Oh I’m sure. Absolutely.

00:22:39:23 – 00:23:10:05
And then because honestly what we’re cause in a new product that we are in the process of getting ready to launch in 2020 we’re actually allowing employees to a more control of their part of their career. And so now where right now we just do we do the reporting to the managers. Now we’re going to be turning it and allowing the employees to take more control of what there’s what they see.

00:23:10:07 – 00:23:41:19
So all the stuff that we’re going through and we’re reporting through the manager the employee is going to have access to as well. So really what we’re trying to do is for us data and policy and whether you’re on the management side are you on the employee side. Once you see even as employee and you see how all of these different things that you’re doing is potentially correct affecting your career progression. This gives you the opportunity to become more involved. It gives you and it helps you along your employee during

00:23:43:09 – 00:23:51:15
Fantastico again. Yeah. So we just take a different view of it. Not to use it for ego but actually use it for good. Oh.

00:23:52:19 – 00:24:30:16
So I’m going to take the last word here and suggest what I described as an evil with all what I describe is what human beings do when given the technology and so design of technology is evolving to account for the fact that people will use the technology in other ways than the designers intended. And it’s a very challenging part of the question and that’s where all of the ethical issues in a A.I. bias and visibility and surveillance and privacy all of those ethical issues end up suffering this very single topic.

00:24:30:28 – 00:25:06:21
And it’s going to be where people differentiate themselves in the market coming right because one of the problems with a is is unintended consequences and we just have this conversation that’s been amazing. We just uncovered a really interesting set of things about unintended consequences that you might actually be able to foresee. And so no. So right. So what a great conversation we’ve run through the time slot. So what are the two or three things you’d like somebody to remember from the conversation that they wish to us or their car going down the highway.

00:25:07:20 – 00:25:37:11
Well I think the most important thing is making sure that your systems are in good working order in your systems are working for you because too often you’re sending all of your time doing stuff outside of the system because the systems really aren’t configured correctly. So there’s that and they’re not up with that you need a way to be able to report and you need it real time. So that executives or even managers can make decisions to adapt. And that’s exactly what they talked about.

00:25:37:14 – 00:25:42:10
Fantastic. So please reintroduce yourself and tell people how they might get a hold of you.

00:25:42:10 – 00:25:58:22
Sure. So my name is Jamie Troiano, I am the founder and CEO of PredictiveHR. You can visit our Web site find out a little bit more about what we do predictive H.R. dot com or in if you need to. You can reach out to me at

00:26:01:17 – 00:26:02:08
Thanks Jamie.

00:26:02:09 – 00:26:11:12
It was a great conversation. I can’t tell you how much I appreciate your forthrightness in the conversation because it’s it’s refreshing, it’s really really refreshing.

00:26:12:08 – 00:26:14:22
Thank you. Thank you very much. I appreciate it.

00:26:15:06 – 00:26:33:02
Yeah. So you’ve been listening HRExaminer’s Executive Conversations and we’ve been talking with Jamie Troiano and you slell that, T-R-O-I-A-N-O, who is the CEO and founder of PredictiveHR and you might want to give them a look. Thanks very much for listening in.

00:26:33:03 – 00:26:35:07
We will talk to you next week.

00:26:35:07 – 00:26:49:19
Bye bye now.


Read previous post:
HR Tech: What Puts the Intelligence in Intelligent Software?

“It is easy to feel overwhelmed by emerging technology of any kind. The following is intended to give you a...