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: Danny Abdo, VP of Solution Engineering, Degreed
Episode: 319
Air Date: March 22, 2019

 

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.

SPEAKERS
Host: John Sumser, HRExaminer
Guest: Danny Abdo, VP of Solution Engineering, Degreed

Full Transcript with timecode

00:00:13:16 – 00:00:25:17
Good morning and welcome to HRExaminer’s Executive Conversations. Today we’re going to be talking with Danny Abdo who is the head of solutions engineering at Degreed dot com.

00:00:26:09 – 00:00:27:11
Danny, how are you?

00:00:27:24 – 00:00:29:16
Doing great. Nice meeting with you.

00:00:29:28 – 00:00:33:27
Thanks. Would you take a moment and introduce yourself, tell people how you got here?

00:00:33:27 – 00:00:36:23
Yeah sure I would be happy to. So, Danny Abdo. VP solutions engineering.

00:00:40:03 – 00:02:13:22
I came kind of stepping back about 15 15 years coming out of college came in to industry as a developer, computer scientist background. That’s I started with an internship at PricewaterhouseCoopers. Where you know before I think we even put the E on learning they were experimenting with some different things they asked the software. I think they sold it probably made a lot of money and it hired me into the learning group which became the learning technology group and I really never left. I’ve always enjoyed education human performance and how we can use technology to make those things better. And I kind of grew up in that career prior to degreed a skipping a few years I led enterprise learning technology at Bank of America. You know where I sort of reached a point where. I had always believed in the promise of technology that this would make things better. And when I sort of reflected that and started thinking about what technology was doing for education you know we made it to make it more efficient. We managed to make it more scalable. But I don’t necessarily know if we we took a fourth step in creating opportunity and actually improving that education and that’s sort of where I met agreed. And I realized quickly they were doing something different. They were really focused on using technology to improve opportunity. And technology had caught up to the place where it could do that. And so I decided to join degreed and I joined them. They were very small at the time. Maybe about 40 employees. And it was it was just an amazing opportunity. I came in for engineering kind of circulated around a number of different departments and here I am today as V.P. of solutions engineering.

00:02:14:04 – 00:02:25:15
So what is degreed do chores. So I would summarize when when when you think about degrees we gave organizations and individuals a better way to build measure and signal

00:02:25:21 – 00:03:15:11
Their skills to the world. We believe that the market wants to speak the language of skills and we’re introducing the ability to do so with with degrees. We sweet we are mission driven company. We started out with the mission of you may have heard it if you’re familiar with the degree with jailbreaking the degree and it wasn’t to say there’s anything wrong with the degree it’s just the world has sort of shifted and there needs to be I guess more ways of communicating your expertise you know in degreed realize that and I think the world realizes that they really don’t care how an individual is become an expert just that you are in degree it really enables an individual to become an expert using whatever path they would like and then allowing them to be able to signal their expertise in a way that the market understands. That’s my bit by best summary of degree.

00:03:16:18 – 00:03:26:09
That’s an interesting concept. So at the heart of degreed is the idea that people should be an expert does that right. I hadn’t thought of it that way before. Exactly

00:03:26:09 – 00:03:50:12
Exactly that you know whatever your path to becoming an expert degree wants to help facilitate that path and then allow you know basically to be able to validate that you then have become an expert. Think along the lines of a credential in a way that you know corporations and in the world understand you know that credential as as you being an expert. That’s exactly right

00:03:51:15 – 00:04:23:22
So in order to far down this rabbit hole. But but it seems we’re coming into a time where what we need is more generalists. Right. That that we’ve sort of gotten to a in gotten ourselves culturally into a difficult position where we’re so specialized that nobody to sit back and ask ethical questions. And so so I wonder if there is a path inside of degree generalist for the somebody who doesn’t specialize or

00:04:24:03 – 00:05:22:29
Specializes in not specializing I think so and if I if I’m sort of interpreting it correctly you know what we and we have a lot of data data to show that a lot of the sort of the skills that are needed today regardless if you’re an engineer or a specialist are some of those more general. I don’t I don’t like the word soft skills but but to a certain degree those sort of generalist you know a bill at communicating with impact here and in degree does introduce you know we are exactly able to understand not only if they if they have that skill but at what level you know that that’s skill that skill is. And so there is definitely that that ability is absolutely does not need to be specialized sort of technology or role specific. It can be sort of more horizontal if you know those those general skills you need to just just perform at any role really.

00:05:22:29 – 00:05:55:18
Sometimes I’d like to have a much longer conversation with you. It’s interesting the notion that. Everybody has an expertise that they will communicate that’s further out through the our thought degreed Rose officially. How does a degree do this to help people articulate their skills. What does that mean. Do you have to take your test so that you could prove you do so undermining certification process documentation process yeah.

00:05:55:22 – 00:07:06:27
Absolutely. So there is a process and this is this is something that’s you know proprietary degrees that we spent a lot of time building. But you know most assessments are focused on what you know and test what you know. Not many are focused on or if any on on the actual outcome. Right. And so our assessments built around outcomes can you actually perform this skill at a specific level. And so you know there is there’s about seven different inputs that go into this. This is a think about it maybe akin to a short application to getting accepted into grad school. Let’s let’s stay. But we we asked for to submit evidence we we then you know take it’s reviewed sort of socially and then reviewed by experts at a summary level without going too deep. And those are sort of the major components that then allow us to effectively assign a sort of level around a specific skill for for an individual who is more oriented towards white people than blue collar people and to person talking like we work

00:07:07:14 – 00:07:25:23
With people who do vast bulk hiring you know companies with half a million or a million employees doing the same job in a lot of ways. So the distinctions between people and their jobs are not particularly extraordinary. Does degree apply in the world.

00:07:25:24 – 00:08:22:26
No I think that’s that’s sort of the value in it. You know to be sort of universal and to run this process regardless of sort of what that what that skill is you know with its outcome base and it’s requiring evidence you know it could be performing more of a blue blue blue collar type of skill right demonstrating that skill. I don’t I think that is sort of what what makes it awful. And we need a universal way of doing that to understand you know across all skills you know and understand that you know a universal leveling system. And I think once that has to be established for there to actually ought to be a true currency and for there to be real value in it you know that that has to be the case and that is our aim our goal and in what we’re doing for organization shows however is spend one more second here then we’ll move on to the rest of the conversation.

00:08:22:27 – 00:09:28:14
What are the things that I’ve learned about skills is that it’s also the case that a job just to a resume resumé glosses over the most important skill with a job. And so the classic example here two nurses giving shots and if you look at reservations they never mentioned the fact that nurses give shots and if you look at job descriptions they never ask for nurses to give a shit. But if you want to if you want to predict the likelihood that a nurse is going to be successful it will be the intersection of his ability to give a shot and the number of shots that are required to be give within the specific role. And so you don’t want somebody who gets sued every time they use a needle in the flu clinic you know. And so so does the feed have that level of nuance that’s missing from the rest of the job market in skills.

00:09:28:14 – 00:11:21:11
Yeah I think so so. So if you think about let’s just add we’ve got a lot of organizations. You know health care organizations that use the platform. So if you think about the inputs that are coming into help understand what skills a specific roles need the degree to taking advantage is taking advantage of a number of these. One is the organization has the ability if they have it and not all organizations know what the specific skills are most critical for roles. If they do we we take that as an input. We are also looking because we have a number of different organizations right across industry verticals. We can we can help other organizations that might not have that data to understand what skills similar organizations are are associating with that role. And then we’re also able to look at the individual is another input skills the inputs that they’re giving into that role what they think they need. Right. These are these are three data sets that are sort of coming in and then we’re able you know when you sort of start getting into that data science and how we use that data to to to predict in you know great accuracy what are the most important skills for that that role and then the beauty is right that it can be dynamic that can change and that’s what that’s what we’re really starting to see with a lot of roles. You know maybe not a nurse but these other roles you know the skills that are required are evolving at such a great pace and so this ability has to work in in sort of real time and update these roles as that you know block chain and all of these new concepts enter in. We know these roles can’t be these roles and skill associations can’t be static. And so so part of what we’re able to do is is update these in real time as the market demands new skills for you know traditional roles. So

00:11:21:11 – 00:11:23:26
So if I’m an employee.

00:11:24:20 – 00:13:01:08
Cary Grant to spend more time keeping my pro forma Well we’ll know so that’s that’s that’s part of. So if you’re using sort of the way I think that’s the wheel would turn for an individual that’s using degreed is they may they may come into the platform for the first time. No platform understands what their current role is if they’re using it as an organization helps them understand what are the specific skills needed you know degree it may prompt you to do just a quick assessment to understand you know baseline where you think you are versus what the target is. That right there you know we could properly type deep up those two data points are extremely valuable right there. But but based on that we’re now able to say OK there’s there’s sort of a gap here. This is we’re learning and enters the equation and knowledge right. How do we then guide that individual to close close that skill gap and then once once we’ve seen the activity and we’ve seen the user sort of progressing in that cycle we can prompt them to do that sort of skill verification or validation survey that helps then you know give them and the organization a real sort of a signal with real veracity that they say that they’ve acquired that skill. And so it’s very continuous and it’s very it doesn’t it doesn’t ask or require too much from the individual. It sort of happens continuously because employees are developing maybe skills that they’re interested in and for a new career for their specific role and so it’s just it’s sort of built into that experience if you will.

00:13:01:08 – 00:13:09:19
So what goes through natural language processing or neural networks play the way that the group works.

00:13:11:12 – 00:14:55:13
Yes absolutely so it’s a great question and we definitely rely or utilize machine learning to it to a great degree. And because it’s kind of using that cycle that I just mentioned right. What we’re what we’re finding is an organization to an employee’s first and foremost might not even know this the specific skills they need to achieve achieve their goals. Right. So that’s that’s sort of the first the first piece and so if you kind of play that out if we’ve got you know four or five million people using this platform and in their entering in you know skills that they’re developing and we’re seeing what types of learning they’re using to develop those skills we’re able to harness that for the entire community on degreed and be able to first and foremost feel the helping individual understand what feels they need to develop today and then it starts to get into you know being a little bit more predict you know once and then being able to help them crosswalk with learning predict the learning that they need to close a skill gap and then what skills do you need net and become predictive and so I think using whether it’s using skill data or it’s using learning data. Our goal is to understand these associations across our user base and understand this activity and be very predictive and really give value to that to the individual the organization using this use using our data and helping them predict what skills are going to need next. And then also help them crosswalk and develop those skills but with learning across the learning ecosystem.

00:14:55:19 – 00:15:01:21
So give me a sense of how you think you see the predicting which schools are given.

00:15:01:21 – 00:15:22:02
The next thing I think this is above the boob gets a lot of people learning and the robe and tangled up with knots. And so if you go up if you’ve got some juice that you can apply to the do long term skills for transferring problem I’m all ears Sure.

00:15:22:04 – 00:16:58:27
Sure. Well so. So let’s split my time using an example. Is this the best way to maybe be articulated. Absolutely yes. So let’s let’s take that let’s say OK we’ve got four million four million users creating a huge data set on degrade this relationship of what what. That they’re adding to their profile what learning they’re taking to develop those skills. So now we’re able to say let’s take a trend across 5000 people in this specific role agreed it’s just noted that 70 percent of them have added this new skill to their profile. They felt that they needed to develop that that skill. That’s that’s a that’s a sort of a signal right. We’ve seen we’ve seen that. And then out of out of those 5000 people you know 40 percent of those people are using this learning in the in the ecosystem to develop this to help them develop this field. So pervasive basically what we’ve we’ve done there is you know the organization at that point may not have even known if this goes back to sort of empowering the user. And sometimes they use or the employee might know best because they’re there on the ground maybe with their manager and they realize you know in order to achieve this goal I’m going to have to scale up here that where we’re capturing that data and we’re able to surface it you know aggregate and surface that to the organization before they might even know they need that skill for that for that specific role for that show maybe maybe one example.

00:16:58:27 – 00:17:48:00
Do you have do you have a method for correcting this. Bruce ruse in 1999. If you were to do what you just described then you would be recommending to everybody who shoots or recruiting just a few moves for cram because the Y2K problem of everybody tied up in the knot and lots and lots and lots of people learn fourth term because the Y2K problem was going to be so severe. And then it wasn’t right then. So this was the problem in general with machine learning is that it predicts what to do with the herd isn’t always right to watch the correction now.

00:17:48:01 – 00:18:26:21
So it’s a great point. We spend a lot of data science team spends a lot a lot of time here right. So I think you can on a really really important point. And so what I what I often say is you know the right blend here is there is a combination there has to be a combination of human intelligence plus machine intelligence. You can’t rely especially if you know in the business that we’re in solely on data in data science and so I’ll say you know in a lot of the processes that we use there is a human component.

00:18:26:24 – 00:18:54:29
And even in our own data set we do spend a lot of time examining those those data sets because you could have this even you know you can get it into natural biases you know based on based on your data set. And I think what you hit upon is this whole whole idea of you know sort of that bubble creating these bubbles right. I think you see a lot of it in sort of the news networks and the social platforms that people people use.

00:18:54:29 – 00:19:41:08
They’re just they’re just getting you know recommended and it’s a cycle of just saying what what is in your network and what’s happening in your network. And so there are there are techniques in data science and then in injecting human and human intelligence in sort of examining sort of data set clean that up into help identify where that could be happening. So I think probably the way I was summarized is you still have to have that human element in there. And this also kind of gets into the content aspect of it too and the quality of content in what’s being recommended there is there’s still a human component that has to be put into that process where you’re using machine learning.

00:19:41:08 – 00:20:02:28
One of the really big questions is get your group. We’ve talked a little bit about terms leaving school to do something abuse understandable and when we were doing interviews show you used the phrase skills currency so I’m guessing movements remember the big questions that know you know I think.

00:20:03:09 – 00:21:59:17
I think that’s the biggest piece right is is getting the world to a point a point where I ask you about your education what you know and what you can do that you don’t have to rely on your you know your college education being out I mean it’s really if you think about it’s very articulate if you like me asking you if if you’re in good health and in good shape and you you say you ran a marathon 15 years ago I think what what we’re seeing is there’s you know the way the world has shifted as people are acquiring skills that there’s an abundance of education out there. People are taking advantage of it they’re they’re learning at increasingly fast fast rate and what they know and what they can do is so different than what they were formally educated in. And I think that that is one of the big problems that we’re that we’re trying to solve. So they have a way to reflect what they actually can do you know in in real time. That that drives a lot of this. And and it applies. You know I think it applies back to you know the more nuanced questions which are how do we know then how do we help individuals discover you know that this skills once we once we can do that then we can unlock these other things which is upskilling people in helping connect you know people you know the supply side of it with the demand the organizations that that need skills because now an organization that uses degrees to understand that this skill at this level performed very well for this role and when they’re looking to the market for that day and people with those types of credentials they can find them and you know it’s just bringing the supply demand side together in a really powerful way.

00:21:59:24 – 00:22:02:19
What are the ethical issues of the work.

00:22:02:20 – 00:22:20:12
You know we we’ve got there are a number of sort of ethical issues that we talked about sort of kind of use a few examples here. You know I think when you’re collecting a lot of data there’s there’s some that are very easy to solve. There’s there’s some that become very very tricky to solve right. So

00:22:20:12 – 00:22:54:00
So when you’re collecting tons of data that data have sort of inherent value. And I think that the first thing that probably comes to mind for a lot of people is that that that data can be modified right. There are people that want that data that are willing to pay for this data that that lead is capturing. So that’s that’s an easy one right. That’s just that is a policy we don’t. We just don’t do it right. We don’t. We don’t sell data. We don’t transmit data. We know this data is owned by our client by our by our users.

00:22:54:05 – 00:25:59:15
Another another piece of that is what content when you think about degree being you know we are also a content marketplace because once an individual has decided what skill they want to build we open up the world of content to them. And you know a lot of these content vendors have a vested interest in their own content. You know that’s that’s where they may make their money and so we are agnostic. We need to continue to be fair and balanced across content vendors and there’s they’re sort of you know I guess Google’s probably another example they probably deal with this is making sure that that system cannot be gamed that that we don’t sort of you know help game that system. So making sure it’s very balanced across our content vendors meaning the best content wins and just the best content wins. So those are those are sort of the easier ones for us to solve. I think where it gets more tricky is you know we are providing a very strong value proposition for organizations and that’s why they pay for paper degree but we also have a strong value proposition for the learners which is which is why they engage with degree right things they own even if they’re learning within their organization they still maintain ownership of their profile their lifelong learning and skills in achievement. And so you know they they get value by using that maybe with their current employer but they may they may use that you know beyond. Right. And so it’s that value proposition that gets employees to engage with our products which creates the data that provides value to the organizations and so. So it’s basically a fine balance that we have to strike between our responsibility to to the look to the employee the user and our responsibility to those to the organizations. So really how do we walk this line provide between providing our clients the data they need to a lot of the value that they’re paying for but without breaching the trust of our of our users because if we do that then the whole system breaks down right. And they’re not likely to engage in and so that is that becomes a little bit trickier. So for example you know to use an example if we know you know we or we do know right is that there could be a spike across a whole population maybe in a specific department maybe in a specific role looking you know that that have started developing for a new role and may we know that they might be planning a move before you know their manager knows where their organization knows. And this is an insight that that the organization needs. But we need to be able to provide that insight in a way that the user could not be negatively impacted by if it cannot be targeted by the organization. And so so we do you know it makes providing insights and data that much more complicated because we’ve got to be very thoughtful on how we’re packaging data at what level what we’re providing and cannot be misused in a way that could breach trust either way.

00:26:00:05 – 00:26:22:27
That’s that’s probably you know one area where we spend a lot of time making sure that that chewing a good job in so trying to ask big questions and giving them to a small space for those things that happens when you characterize people in the way we treat categorize categorizes people.

00:26:22:29 – 00:27:22:27
So the lines between categories firm beauty and nuance. Mike Mike Mike happened music room digital the early stages of her incredible picks losers with new CD Europe and over time Berger removed those two the rose. There was a similar problem. Where were we. The subtleties between turnover gurus are kind of trumped to make you smoother the consequence of that is that people’s new skills get trampled on a little bit. Do you have the same thoughts about them. Well this is is is the Ralph pigeonholing of girls. That’s inevitably going to get better. Is there is there a human consequence to them.

00:27:24:17 – 00:27:27:06
That’s a great a great point.

00:27:27:06 – 00:30:08:26
You know I think there is I think about it from too I can I think about it from both both angles and in there are. It is it is something that that is considered but where I would start writing is where we are today. When you when we when we’re speaking with organizations and sort of understanding the landscape there’s just no. There’s such a lack there of of of insight right into into skills in their capabilities. And so we’re almost starting from from zero to a certain degree and so I think I think what you’re saying. So I think there’s is there a huge arc of of improvement that that is just going to provide value. Right it is just going to provide immense value. And if I understand the question though is like then you sort of start plateauing potentially in where you’re starting to lose some some of those those those different nuances. And and I think what we’re what we what we strive and try to do again going back to not just relying on sort of data in any big capabilities is how many different inputs are you able to to process to keep to keep things very individualized and very very unique to that individual. So it can’t just be you know what what skill has has been added right. It can’t just just be level. You know back in the 12th it you know 24 bit you we’ve got to we’ve got to aim to be higher higher fidelity and that’s why we allow. Right we allow when somebody is sort of crafting their profile of who they are and what they can do you know ultimately they have they have a certain control and responsibility over it were worth filling that in with inputs from their organization from from the external social network from their you know from from their colleagues from their manager. And I think the more inputs we’re able to take into that and that but then allow the user that to sort of personalize and reflect what what they know I think that’s one way we we deal with it. But I would like it without going too deep I think the where we’re at now we’re just we’re we’re so early there’s just so much value because we just don’t have much data there. There’s just not not many signals or data there. So I think that that issue over time and I think you know hopefully technology will improve over time as well that that’s the better deal with it. But I think we got quite a ways built that we sort of get that that sort of plateau that we’re going to experience.

00:30:09:00 – 00:30:20:18
Was a in question. So we’ve blown through our time one of the museum conversations. Thank you so much for doing this. Would you take a moment to reduce yourself again and tell people how we might get a hold of you?

00:30:21:14 – 00:30:44:18
Yeah sure. Just to recap. Name’s Danny Abdo vice president of solutions engineering at degreed My, my email address if anybody found any of this interesting it’s a topic that that I would love to continue dialogue around. My email is Danny, just D-A-N-N-Y at degreed dot com. So feel free to email me at any time. And thank you John for the time today.

00:30:44:28 – 00:31:01:10
Yes, thanks, this has been a great conversation. You’ve been listening to HRExaminer’s Executive Conversations and we’ve been talking with Danny Abdo who is the vice president of solutions engineering at Degreed. Thanks for tuning in and we will see you here same time next week. Bye now.