IVP’s Future Of Work Podcast: Eightfold.ai CEO & Co-Founder Ashutosh Garg on Technology’s Role in Finding, Hiring, and Retaining Talent
Welcome to a special edition of IVP’s Future of Work Podcast. In this series, IVP investors talk with experts from the fastest-growing companies and discuss the ins-and-outs of the future of work in the ever-changing environment.
In our last episode of a three-part future of work series, IVP Partner, Cack Wilhelm, talks with Eightfold.ai CEO and co-founder, Ashutosh Garg, about technology’s influence on the future of hiring.
Ashutosh is a founder and former CTO of BloomReach and has held research scientist roles at Compaq, IBM, and Google. Over the years, he’s produced over 6000 research citations, 50 patents, 35 peer-reviewed research publications, and claimed the Outstanding Ph.D. Thesis Award from UIUC for Machine Learning. He’s also a published author of the book “What’s Next for You — The Eightfold Path To Transforming The Way We Hire And Manage Talent.” Quite simply, he’s a preeminent expert in machine learning.
IVP invested in Eightfold’s Series C in April of 2019 to help companies improve talent management, recruiting, and retention — achieving diversity goals and improving employment accuracy at scale. Eightfold is founded on the mission of empowering organizations to transform how they hire, train, and retain a diverse workforce.
Here are a few podcast highlights to help company leaders with talent management:
Shift From Employer to Employee-Centric “If you focus on employees, you will do better as a company than if you just focus on the company, not the employees.”
In the past twenty years, the world has materially changed, given technology’s influence on communication and interaction. This shift has also changed employment, with employees switching jobs every two or three years and careers every seven or eight years. To stay successful, companies have moved away from treating employees as a resource to an actual human with goals, aspirations, and motivations. This shift creates respect, retention, and nods to the changing workforce, underlining key learning: if you understand people, put them in the right place, and allow them room to grow, they thrive — and so does the company.
Recognize Bias in Hiring and Take Steps to Address It “One of the biggest differences between machines and humans is the ability to forget and ignore.”
Strategically using AI in business operations comes with an inherent ethical responsibility. Eightfold understands this responsibility and takes a multifaceted approach to address it. Using career data from more than one billion people, the company is ensuring they’re using a large enough data set. They then use equal parity algorithms to make sure that, at every step of the process, the behavior of algorithms is independent of people’s characteristics like race, gender, ethnicity, age, and other characteristics. With enough audit and monitoring processes in place, a transparent system emerges with functionality, like masking, to improve results. Eightfold puts all of these levers into place to measure how a candidate moves through the hiring process, while significantly reducing the potential for bias in the hiring process.
Eightfold.ai Creates Talent Exchange for Those Impacted by COVID-19
Amid the mass layoffs driven by the pandemic, Eightfold is connecting displaced workers with the right jobs. Talent Exchange is a nationwide marketplace connecting HR and business leaders to immediately employ associates whose jobs are impacted by the COVID-19 pandemic. It understands skills and job requirements on a large scale. Eightfold, in partnership with The Food Industry Association (FMI.org) and supported by McKinsey & Company, is providing the platform free of charge.
More from the full conversation in the transcript is below.
To hear more, listen on iTunes or SoundCloud.
TRANSCRIPT
Narrator: Welcome to IVP’s Future of Work Podcast. In this series, we talk with experts from the fastest-growing companies and discuss the ins-and-outs of the future of work in the ever-changing environment. If you like what you hear, consider following us on SoundCloud or subscribing to our podcast on iTunes. Thanks, and enjoy the show.
Cack Wilhelm: Welcome back to a special edition podcast on the future of work. In this third episode, I talk with Ashutosh Garg, who’s currently the founder and CEO of IVP-portfolio company Eightfold. He is the second time founder having started BloomReach, and also a former research scientist at Compaq, IBM, and Google.
Cack Wilhelm: Ashutosh, thanks for joining us. Let me quickly give your bio. So you graduated from IIT, Delhi, and you’re in the IIT mafia, followed by a Ph.D. at Champaign-Urbana, Illinois. You then held a number of illustrious research positions, first at Compaq, then IBM, then Google. Now Eightfold is your second startup. You’re originally a founder and CTO at BloomReach, where you stayed for almost nine years, and most recently the founder and CEO at Eightfold. Let’s do this; if you could take a minute and explain Eightfold in your words, I think that would be helpful for our audience.
Ashutosh Garg: Eightfold is a company focused on the mission of providing the right career to everyone in the world. We started the company three years back, and today we have built a talent intelligence platform that is being used by some of the largest employers in the world to hire, engage, retain, and build a diverse workforce.
Cack Wilhelm: Okay, great. So, I mean, let’s start broadly. I think Eightfold really emphasizes the early stages of talent management, starting with hiring, all the way through to employee management, and then the ongoing experience once you’re hired. Another company in our portfolio, Humu, who’s also on this podcast, is focused heavily on taking what is the existing talent base and then creating that stronger purpose and productivity within the workforce. What that tells me is, you know, we have two companies, both loosely in HR., tackling very different problems, and it says there’s got to be a ton of opportunity to evolve HR, human resources, as we know it. So what do you see as an opportunity in HR broadly?
Ashutosh Garg: So if you think of HR as space, right? Over the last 20 years, the world has changed materially. Few examples of that: one is the rate at which these skills are evolving is faster than ever. In the past, you are working on the same thing for years and years. And today, every few years with technology stack completely changing, the way things are sold changes, the way people communicate changes. So that’s one big change that is happening in our society. The second thing is, in four years, 75 percent of the workforce, globally, will be millennials. What that means is how we think of talent can no longer be the same thing. People care about their careers.
Cack Wilhelm: Yeah.
Ashutosh Garg: It’s no longer about jobs. People are switching jobs every two, three years. In fact, many people are switching careers every seven, eight years. Most people will have at least three careers in their lifetime. But on the other side, your manager has been in the same job for less than you, in general. So no one is there to guide you with what you should be doing. No one is out there for an employee to handhold them and guide them in this process. On the other side is an enterprise. Talent is more important to you today than it has ever been in the past. The whole asset of Google, Facebook, Salesforce, is the talent they have. So how can they stay ahead of the curve? The only thing it depends on and has, is talent, right? Now, today’s HR infrastructure that exists in enterprises, it’s from twenty years, now, that the infrastructure that has not changed one bit. Today, if we focus on the whole enterprise experience of payroll and benefits on one side. On the other side, you still focus on applicant tech, info, something that is tracking the candidates. But if you go and ask any large enterprise, how many people in your company can really learn blockchain quickly? The answer is — you have no idea.
Cack Wilhelm: Mhm.
Ashutosh Garg: How many people in your company can be put on this project based on their current and past experience? The answer is — you don’t know that. So the HR has not evolved. At Eightfold, what we realized is that there is a huge space of talent that is untapped. Outside the OHRS and APS, enterprises need somebody to better understand, manage talent, and not manage in the traditional sense of payroll benefits, but manage talent in terms of professional potential, and so on. So straight when we are going and talking to enterprises, right? We position it as an entire talent stack set starting with, who you should be engaging, how you should be engaging, who you should be bringing on board in your company, how you should be allocating projects across these people as they are coming in. Who you should be upscaling, who you should re-skilling. What skills you should be focused on, what skills your competitors are focused on, what are the skills best of the breed companies are focused on, and enabling them to achieve that.
Cack Wilhelm: Yeah, that’s really interesting. One thing I noted is you talk a lot about the shift from being employer centric to employee-centric. And my guess is that it is a big adjustment for enterprises. I mean, can I be flippant and ask, is this driven by your comment about the emergence of the millennial for the workforce and what percentage of the workforce they’ll make up? And with that, you know, enterprises, I imagine it has to come full circle to where enterprises see the benefit of operating this way or they just wouldn’t adopt. So what can you reflect on that?
Ashutosh Garg: Your question is why we talk about being employee-centric versus employer centric?
Cack Wilhelm: Mhm.
Ashutosh Garg: There are various ways to look at that. Once you understand people and you put them in the right place where we can thrive and grow. The outcome you get from those people is so much more than otherwise. It comes from our experience, right, anything we have the flexibility to execute in our way, we have done so much better than otherwise. So the goal is not to say that it is about employee or employer. Employment is really bringing the two together in the right session. And in the past, humans have been treated like a resource, human capital, human resource. It’s not about just be the one body. These are people who have aspirations, right? If you can guide them, motivate them, then they will do wonders for you.
Cack Wilhelm: Mhm.
Ashutosh Garg: So the goal is the same to make enterprises successful. But everyone is now realizing that if you were to focus on people, then they will do so much better otherwise. And this is what I saw at Google. I think Google did a phenomenal job when it comes to people, talent in their early days. They were like let’s bring the best people and then give them an environment where they can thrive, let them innovate. And personally, I had an even more unique experience. When I joined Google, I joined the research team; I didn’t even meet my manager for the first two months. All I was told is Ashu, now that you are here, go figure out something to help the company. So I think it is more driven by if you focus on employees, you will do better as a company than if you just focus on the company, not the employees.
Cack Wilhelm: Yeah, that’s exciting, thinking about that as the future state of work and engagement with an employer. On the people side, hiring has a lot of similarities to venture capital. So let’s take a customer on your website like Capital One. You know, they have to find candidates using Eightfold, you know, for us, we have to find companies, then vet. But then there’s a whole piece of selling or closing the candidate, in our case, selling or closing you on the idea of working with IVP to come work for Capital One. So you guys can very clearly solve the workflow and sort of the A.I. augmented vetting and all of that. What about the interpersonal, the selling, winning, closing, part of hiring? Do you expect this to remain very focused on human interaction? And also, have you learned best practices from your vantage at Eightfold on how teams can do better at this winning piece?
Ashutosh Garg: Absolutely. That is a very, very interesting and a good question. Today, the process is, we go reach out for candidates, largely in a very cold email saying that you have an opening for this role, please come and engage with us. And the person shows up on that website, reads the job description, which could be full of buzz words, or too lengthy, too many requirements, and most people won’t even apply. If they decide to engage, then they will go through an interview process, four people will show up on the interview panel, they will give us feedback, companies make an offer if everything goes well, and the recruiter or hiring manager then will be involved too with the candidate. On the other side, what does venture funds do? And they do a phenomenal job at this thing, right? First, in most cases, funding does not happen in one meeting. You have to build a relationship with those people. Second is, they try to connect with you in a human fashion, and they try to communicate information that is relevant to that individual. And they are involved in that process all the time. So the A.I. and automation, the goal of that is not to take the human away from the process, or to make the process more mechanical, but the goal of that is to actually make the entire process a lot more human. And let me explain how it manifests itself. So what we have done as an example, we now follow the career website for many of our customers. When an individual goes to that career website, the system says that instead of you start searching for the jobs and trying to figure out our terminology internally within the company, just go give us your profile, and we can tell you what the most relevant job for you is. Now, what this does is that it speeds up the process dramatically. Now, as a candidate, you are not scanning through 500 job descriptions, but now you see the three that are most relevant to you. But then, there are not only these three jobs; it starts explaining to the candidate why this is a good job for you. You’re very likely to get it, the other people who are in this job are very much like you. They have the same set of skills, but then also highlights a few people this person is going to work with because, for us, I think ultimately is it’s a social world. You want to be able to relate to those people. So we now start exposing the employees of the company to this person, who this person is likely to work with, making it a lot more human. If this person does not know a few things, they may be able to communicate both to the recruiter and the candidate that you might not know these things. But that’s OK, based on everything that you have done in the past, we are very confident you can learn that stuff quickly. And, each of these things are solving for different things. Some people don’t apply because they are afraid of failure. Some people walk out part of the process because they feel they will not be able to succeed in that environment. Some people don’t accept the offer because they’re like, “I’m too good, I don’t know who I’m going to do work with or if it will be good or not.” And what it does is that now it starts to answers all those three questions for these people. And as a result, what we now see is that the number of female applicants has gone up by more than 65% in our customer base. And the reason is because like there is this Harvard study, which showed that women would not apply for a job unless they feel they are a hundred percent qualified, right? Now what we have done is, we have taken that assessment out of the picture and said that she would qualify, you know, she is a good fit, you know. You have all the skills needed for the job, don’t worry about it. And suddenly now a lot more of them are applying. Now what we are also helping enterprises understand is are there people in our company who might know this person and not just because they’re connected on LinkedIn. But most of us, they might have worked with this individual in a previous company, they might have gone to the school with this individual, so those who took it further help in choosing a candidate.
Cack Wilhelm: Huh, smart, so you can actually have warm intros where you might not have realized you had them.
Ashutosh Garg: Exactly.
Cack Wilhelm: I mean, a lot of what you just talked about is sort of the positives of the A.I. enabled A.I. algorithms, how you provide leverage to people while maintaining human. Can we just touch on, I mean, there are probably some negative impacts, like what if there’s bias in the training data or the underlying algorithm, or what responsibility do you think a software vendor like Eightfold has to expose the logic leading to a hiring outcome or a decision versus just simply serving an answer?
Ashutosh Garg: So let me start by one other question is. Many things, when you ask individuals what the biggest difference between a machine and humans and the standard answer is more compute, machines can do more processing. Disks can have a bigger memory; they can access data a little more quickly. But one of the biggest differences between machines and humans is the ability to forget and ignore. You cannot make humans forget things easily. At this point, you know my candor and you cannot ignore it, but a machine can ignore it and can be made to forget. Now, how does that map to what we do? When you are using A.I., why we can solve for many things, but it also comes with a huge responsibility. And so it is extremely important to ensure that there are no biases, but it’s just not having no biases, how do you help pilot the biases that may even exist in the organization and make it good. Because the goal is not just to remove the bias, but really solve the problem over here, right? So I’ll give a few examples of what we have done, and in other cases, when I’m more scared than otherwise. Here is the powerful technology, which, if done right, can do a miracle, but if people don’t know how the thing works, then they can completely mess it up. And that is what happened in the Amazon case. They changed it wrong; they validated it with the wrong data, they didn’t put enough transparency in the platform, and that led to disaster over there, right? So very first thing you have to ensure is that are you using a large enough data set. So in our case today, we are planning our algorithms not based on just one company’s data or two company’s data, but pretty much entire workforce. So we are analyzing the careers of more than a billion people to build our models. Second is that the algorithm that we are using, we have invented what you call equal parity algorithm that ensures that at every step of the process, the behavior of algorithms is independent of people’s characteristics like race, gender, ethnicity, age, and even unknown characteristics or that the ones that are not really, things like is someone part of LGBT or any other groups, so we do that. The third is you have to put enough audit and monitor processes in place to ensure that your algorithms are doing the right thing. Fourth is, you want to build a system that is very transparent, so that you can be always explained to the hiring manager, to the recruiter, to the candidate why and what decisions are being made, in a simple fashion. Because if I go up to a candidate today, the biggest frustration as a candidate is, can you just tell me why you’re not hiring me? And companies are like, no, no, we can’t just tell you that. On the other hand, if I can tell that to a candidate that you’re not qualified for this job because this job requires these three skills, which based on your profile and experience, we believe you don’t have. But on the other hand, if you believe that you had those skills, can you give us more supporting evidence, so that we can incorporate that? Now suddenly you have been given the opportunity to look into it and understand what we have for the job, what we don’t have for the job, right? And how can they acquire it? At the same time, you can tell a recruiter that you are seeing a candidate. This person has these skills, and these three skills of their’s are required for the job, please interview them on these three skills because this person does not seem to have those but in a very transparent fashion. Other things that we have done through technology now is a functionality around masking. And what it does is from every profile, when a recruiter or hiring manager is looking at the candidate, it can remove things like name and indicators of gender, race, ethnicity, age. It can also mask the name of schools and colleges. And that has dramatically reduced the bias. And then, after that, what we do is we also provide rich analytics so that we can measure how a candidate is moving through the process and is a bias creeping in? Because the extent of pushback could be, sure Ashu, you can remove the name and gender using the screening pen? But when people show up for the interview, they will see it. Absolutely, they will see it. And that is where the analytics comes in that I can say that you know what, you have screened forty women and fifty men, but you only made an offer to two women and ten men, that is a huge bias over there. So we have like for some of our customers the diversity platform has gone up by almost 40 percent as a result of this thing.
Cack Wilhelm: Well, that’s impressive. Here, I thought maybe I’d stump you with that question, but you clearly have thought very thoroughly about this.
Ashutosh Garg: And the other important thing over here and I would just like to add to here is, what problem are you solving for, right? And we always say that the problem that we are solving for is the right career for everyone in the world. And that means we have to solve not only for the people who had but also for the people who need help. And that is why we have a huge focus on diversity.
Cack Wilhelm: Yeah, well, the right career for all the people in the world sounds like a large task ahead of you. We’re excited to be working with you together on it.
Ashutosh Garg: Thank you.
Cack Wilhelm: Thank you to everyone today for listening. And I thoroughly hope you enjoyed our discussions.
Narrator: Thank you for listening to IVP’s Future of Work Podcast. You can learn more about us on IVP.com or join the conversation on Twitter by tweeting @IVP.
Originally published at https://www.ivp.com on April 30, 2020.
