Mihir: Hello, and welcome to the second of the three-part webinar series focused on innovating the candidate experience. My name is Mihir Gandhi, and I am Head of Marketplace Operations at Eightfold.ai. As a hiring manager for nearly two decades, sourcing, hiring and retaining talent has been a central theme in my career managing at a hyper-growth company like Lyft where I was the first general manager for their Flagship region in northern California. I’ve acutely felt the pain of hiring rapidly and hiring right. I’m thrilled at how Eightfold is addressing these challenges and so much more.

Today, we’ll talk about how that’s happening. Specifically, we’re diving into strategies for recruiting and talent acquisition and how AI is changing the recruiting process. We’re joined by our esteemed guest, Amit Prakash. Amit is the Co-Founder and CTO at ThoughtSpot. We’ll have robust content to cover with regards to how he approaches AI, in general, and also specifically talent acquisition and the interview process.

First, a little bit about Eightfold. At Eightfold, we’ve created a talent intelligence platform for enterprises that leverages artificial intelligence to hire, engage and nurture talent. Talent-centric applications are built on this constantly-learning platform, enables enterprises to manage the entire lifecycle from prospects to candidate to alumni. With over 100 companies as paying customers including Tata Communications, AdRoll, Hulu, Grand Rounds, Nutanix and more, Eightfold has helped companies vastly improve their talent acquisition, talent diversity, and talent management capabilities. Historical Legacy products like ATSs were developed to replace the process of tracking paper resumes, and as such provides pretty similar workloads.

Now ubiquitous and onerous online application processes that companies require from applicants is unduly hard on both applicants and companies replacing paper problems with digital ones. Eightfold was born in the AI era specifically to address and solve challenges with employment in today’s society. As this slide communicates, more information than ever is being communicated about jobs, about companies, and about candidates. These reside on job boards, career pages, social profiles, special profiles like GitHub, Dribble and more. Companies have more information than ever in their ATSs, in their HRISs, CRMs, et cetera.

Of course, hiring managers have specific vision as to the skills, experiences, knowledge, and culture that they are building. More data isn’t necessarily better. It just means there’s more places to find dispread information and try to cobble it together to get a more holistic view of a candidate. It’s not humanly possible to take in all of your data and identify if candidates fit or to figure out their potential to excel in a given role let alone their career trajectory. Then to do this across thousands of candidates and hundreds of jobs is simply a superhuman task.

That’s where Eightfold’s talent intelligence platform comes in.

Eightfold was designed to improve the lives of candidates, recruiting, HR, hiring managers, employees, and alumni. The platform aggregates and digests these data marrying internal data like your ATS with a plethora of sourcing and recruiting tools you could be using with externally available information to create an enriched talent repository. The Eightfold platform uses these data to help surface what candidates are good at today and what they’ll be ready for in their next steps in their career. This drives better talent strategy and talent execution. Once the Eightfold platform has ingested robust data from Legacy and public profiles for each person and thus created a rich profile of each candidate, the platform calibrates each role according to the specific needs of the organization. The results impact the entire ecosystem. The intake process is redefined around content. An instant pipeline of highly qualified candidates is delivered to recruiters and hiring managers. The candidate experience is transformed from being, “Do it yourself,” to, “Let us help you.” The internal referral experience is streamlined to be friendly to recruiters, employees, and candidates.

As we talked about in yesterday’s webinar, the platform drives retention is smart and targeted internal mobility. After our discussion with Amit and before our Q&A, Jason Gray, Eightfold’s Director of Sales Engineering, will give us a brief demo to bring these words to light. Additionally, I’d love to have the audience participate by sending questions throughout the course of the webinar that we can then engage with Amit during the Q&A session. I’d like to give a couple of examples of how Eightfold has helped our customers. This first example is from Hulu.

Hulu was using a host of tools to assist in recruiting such as Jobvite, LinkedIn Recruiter, agencies, job boards and sourcing tools. They were receiving more applications than their team could possibly keep up with. Given their hot growth and trajectory, the volume of applicants, the number of tools they were using, the Hulu team found that highly qualified candidates were slipping through the cracks.After implementing Eightfold, which aggregated data across all of these tools and sources, Hulu was able to have a single view on their entire talent network. On average, recruiters saved about four hours per day and quickly stopped using in-mail since their talent pipelines were full of highly qualified candidates.

At Tata Communications, they were similarly overwhelmed with massive inbounds. Hiring processes were in line with a 10k plus company that’s experiencing close of growth. Hiring managers were spending time interviewing candidates that weren’t the exact right fit. With Eightfold, Tata was able to immediately rank, sort and prioritize candidates who were the best fits for each job. Recruiters and hiring managers were able to calibrate needs in real-time driving more efficient hiring manager time-use during their interviews. Fifty percent fewer meetings could recalibrate on expectations for candidates and approximately four to six- hour saved per recruiter, per day. Now, almost two-thirds of Tata’s hires are driven by Eightfold. As I mentioned earlier, Jason will be giving the demo of the Eightfold platform for the Q&A to help bring some of these examples to light.

Now, let’s get into our conversation. I’m very excited to welcome our featured presenter, Amit Prakash. Amit is Cofounder and CTO at ThoughtSpot and has built numerous high-performance machine learning and analytic systems. Prior to ThoughtSpot, Amit led multiple engineering teams in the Google AdSense environment. Prior to that, Amit was one of the early engineers on the Microsoft Bing team where he built a web-scaled graph computing system responsible for computing algorithms and capabilities like page rank on graphs over trillions of data. Amit is also a co-founder of, “Elements of Programming Interviews,” books where he has tried to help both interviewers and interviewees better prepare for technical interviews. He received his Ph.D. in computer engineering from the University of Texas at Austin and a Bachelor of Technology in electrical engineering from the Indian Institute of Technology in Concord. I’m so excited to welcome you today, Amit. Thank you for joining us.

Amit: Thank you so much for a great introduction.

Mihir: Absolutely. Amit, let’s jump in because I think there’s a lot of content to dive into here. You spent several years at Microsoft and at Google, and there was tremendous growth at each of those companies in your time there. Your teams must’ve grown dramatically during that time, and so you’ve had to hire a lot. As a hiring manager, why is hiring so hard?

Amit: I think hiring is hard because everybody’s realized that recruiting is the most important function in a company. If you get the right people in, there’s nothing like intensive growth and success. There’s a lot more demand for great-quality people everywhere you go than there is supply. What you end up doing is essentially playing “Moneyball” with your candidates because you know that the superstars where whom everything looks perfect on their resume, everybody’s after them. You can go after them, and sometimes you’ll win. To really succeed, you need to find your edge as to which dimensions that you can look at, that you can predict somebody’s going to be an awesome star that not everybody else is looking at. I think that’s where most of the recruiting energy needs to go.

Today, it goes so much more into the process and to the mundane aspect of searching through LinkedIn profiles and resumes and just inbound applications. That sort of makes it very hard. At the end of the day, when you find the right candidate, you need to spend a lot of one-on-one time with them to show them your vision of why it’s going to be a successful place for them to come. To get to that point where you know who you’re looking at is going to be a great candidate, there’s so much effort that goes in. There’s such a small percentage of those applicants that actually pan out that you end up wasting a lot of very, very important time. If there was a better system, you could actually concentrate on the right candidates much more.

Mihir: As you say that, the mental image that comes to mind is a really broad funnel, top of funnel that narrows very quickly. That broad, top of funnel as you describe it is a sifting through hundreds if not thousands of profiles. It’s pretty laborious and intensive. As a co-founder of ThoughtSpot and as a hiring manager, can you describe your interaction and engagement throughout that top of funnel piece and how you widdle it down to the few candidates that you’re going after?

Amit: To make it concrete, I can talk about a specific role that I’m right now looking for. I’m looking for somebody who has a good experience with statistical techniques, data science techniques as well as somebody who is an engineer and has exposure to machine learning. This being a hyped up field, almost everybody is writing those buzzwords in their resume. You did not cross a single resume where there’s not a mention of machine learning or a mention of data science or something. The moment you start to funnel, within five minutes you realize that they have maybe used a couple of machine learning tools and have a couple of projects but don’t have nearly the depth that you want. Again, it becomes a very laborious process to go through all these resumes that match the keywords that you’re looking for and then actually figure out who has depth in there or not. Does that answer your question?

Mihir: Yes, it does. I think what you described is the ability to quickly ascertain the reality of someone’s experiences versus maybe the thought of what they purport to do. Can you describe a tool that you’ve used today or in the past that has helped streamline that top of funnel so that your time as a hiring manager is more efficiently spent on good candidates and not on the 25 extra minutes of interviews that you know you’re not going to proceed with?

Amit: Yeah. Unfortunately, I think – I won’t say that it’s a fault problem. For us being a startup, what has happened is that since all other things are extremely noisy, the most reliable signal we have with employee referrals. We’ve been able to find a lot of really high-quality candidates to referrals. At least a year ago, we were not even in a position to invest the human energy needed to go through all of the inbound resumes and things like that and the ignoring that. Now, we realize that there are a lot of high-value candidates in there that we could have gone after. We just didn’t have the manpower to go through it. Referral-based hiring has been extremely helpful to us. We have an amazing team, very, very high-caliber individuals.

One problem with that is diversity because you tend to know people who look like you, who have gone to the same colleges. Those people also know, again, other people who look like you and went to the same colleges. Your entire workflow starts looking very much like the  co-founders. I would have loved to have introduced this to people with diverse backgrounds and give us the confidence that we could go after them.

Mihir: We see something somewhat similar here at Eightfold where when you grow from two to ten to 20 and beyond, each hire is incredibly important and impactful to the organization. That doesn’t change when you’re a 1,000 or 2,000 or 5,000. Generally, in that process, you do lose candidates, really highly qualified candidates because the bandwidth required to actually sort through and sift through them simply doesn’t exist. How have you seen, in your experience, this evolve over time? Can you describe to a recruiter or agency what you’re looking for in a candidate to help drive this forward?

Amit: Yeah. I don’t think we have really innovated there. We still go through a painful process of writing a job description and talking through what is important, what is not. One of the things that one of my mentors told me at Google is that most human beings are good at looking at maybe five or eight dimensions at most. Beyond that, it gets very hard for a human to put all the details in their mind and work through a lot of data. That’s where machines do a better job than humans is when there’s tons of dimensions that you need to care about. A machine is going to be very dumb in terms of depth, but it’s going to be great in terms of breadth. Sorry, I lost my train of thought.

Mihir: No, no. That’s helpful. When you say that, how do you and I think your work at ThoughtSpot is actually very pertinent. You see AI in tools that drive those types of deeper insights, empowering the human element that’s required to do the work.

Amit: Yeah. I love the engineers on my team. I think we have an amazing team. What I tell my recruiting team is that “If you can just find me exactly these kinds of people, just send more of them that will be fantastic.” It’s very subjective to know what do these kinds of people mean? Do they need to have gone to the same colleges? Do they need to have the same job experience? Do they need to have worked on the same project? That’s where a lot of subjectivity creeps in.

Sometimes I see my view of who are the high-value candidates, and the other is different from the recruiting team. Then, that leads to missed opportunities in terms of who’s candidate was prioritized higher? It’s somebody who’s just waiting for a week because you’re rushing through those other candidates.

Mihir: It’s interesting that you say that. You mentioned the job description process and frankly how broken job descriptions are. That is actually the candidate facing a description of what they would be doing. The reality is you’re looking for someone like someone else. You say like someone else. Can you help describe a little bit more about what that means? How do you actually translate that to a recruiter that then translates that to whittling through the top of funnel?

Amit: We had a good success initially hiring engineers with years of experience from Google. That was a great profile for us like people who have spent a few years at Google, the best place to learn engineering on the earth. It’s also very hard to go to people who are in very, very high demand. People like them, again, there’s so many dimensions to look at. It could be the college. It could be the project. It could be the quality of their project. It could be that they are friends with somebody that is in my trusted network. It could be that they have published something that has won awards or something like that in a confidence that I know high-quality talent through. That’s where once you’ve spent a lot of time with your recruiter, they can get to see how you evaluate candidates and get a better and better picture all the time.

Mihir: The dimensions that you outlined are almost impossible to sustain, stack rank and prioritize a school like this or experiences like this or evaluate they’re getting. The reality is candidates are so diverse that the ability to stack and prioritize is a nontrivial task. When you talk about spending a lot of time with your recruiters, you talked a little bit about the recruiting channels that you’ve used. We’ve seen some pros and cons at different types of recruiting channels be it advertising channels, agencies, in-house recruiters and so on and so forth. What’s the level of success across these modalities? What are the pros and cons? How do you spend your time investing your prospects?

Amit: In my personal experience, there’s nothing worse like your team’s referrals. That’s probably the highest priority by a factor of ten or so. Beyond that, I think we haven’t had much success with agencies and such. There are a few areas where it’s a very routine hiring process, where you don’t have to look at many dimensions. You just look at a few dimensions. You’re okay with that. In those cases, we have had decent success with agencies.

The referral effort is great where it doesn’t scale well. The other thing is there are a lot of inbound candidates. Potentially, what we have done is that we have staffed up a recruiting team so that they can go through all of the inbound applications and figure out which ones are worth going after and which ones are not.

Mihir: What you’re describing is a bit of a tipping point where you’re going from being the hunter trying to get people to come into the organization to being an organization that has so much inbound that now you need to staff up, provide the tools and processes and resources to whittle through that. I think one of the things you mentioned earlier is when we were talking before the webinars.

There’s very qualified candidates who simply just fall through the cracks when you have that massive inbound. Can you talk about some of the challenges that you face with sourcing and screening post-tipping point when you’re fielding all the inbound and actually kind of whittle through those folks who are interested in working for you?

Amit: I think when you talk about if you start looking at your inbound, there is a narrow fraction that obviously is good. You can’t know that they’re going to be good because there’s a very small fraction that’s very much in the line. Most of the time that time you reach back, either they have moved on or they have their props and creds. That’s a great pool, but the success rate of that pool is not that great. Then that’s where creative thinking and analytical approach comes in where you find things that are not obvious to the market that are good predictors of the success.

For example, there is a tiny company, not tiny actually but a decent-sized company in India called Directi. We’re hearing as we talk to people, nobody knows about them. What we have found through hiring two or three people is that that was fantastic training them for a lot of great engineers a few years back. Any time we see somebody from Directi, we go after them. Similar things that we have found is that people may not have a great pedigree. If they have participated in programming contests, particularly a few good ones that tends to be a very good indicator of their performance. They need other aspects too, but you can test in the interview. In general, that’s a good signal to go after.

Mihir: It’s so hard to communicate that to a recruiter or to an agency, “If this, then that but not this,” and, “We’ve seen some traction internally around this. Can you go proactive with a better analogy? Go fishing in that pond or in that part of the lake where we have had success in the past, and we think we have competitive advantage?” If you were to take a step back from using modalities, can you talk about what an ideal sourcing process looks like for you?

Amit: I think an ideal sourcing process would exactly look like what I said. I have a team of 500 people. I know that they are fantastic. “Just find me five times more of that.”

Mihir: You almost don’t need to say more than that. I mean you shouldn’t have to say more than that. There’s such rich data that already exists not only in your head but on their resumes and their publicly available information that it should be the extent of which you engage at the top of funnel. When you –

Amit: The other thing that I’ll add to that is that not everybody that looks great is recruitable. The other signal that I would love to be included in the sourcing process is people who are likely to perform. This could be based on how long they’ve been at a company, or there has been communication from their side to the world that they are ready to move on or maybe an additional change indicator. That’s the piece that’s hard. Once you found what looks good and you found the signal that they’re looking to move, that’s when things really start moving.

Mihir: Those are not impossible to find. Those data exist. They’re just difficult to parse. If you take the time to look at someone’s resume over a five or a ten- year career span, you generally get a sense of could they be stagnating or ready to move? It takes a lot of time, a lot of resources to do that and frankly a fair amount of skill. When AI is developed correctly and applied in the right way, how do you see it helping recruiting?

Amit: I think where AI does a fantastic job is when there’s lots more than eight or ten dimensions to consider, and you have some data that lets the machine say, “What correlations have what we are looking for and what correlations have what we are not looking for?” This seems to be a great problem to apply here too. In an ideal world, all these different signals that are available about a candidate would be processed by an AI engine to properly scale and tell us that ‘A’ they are going to have great career potential and ‘B’ they are likely to consider.

Mihir: A lot of that time is currently spent that recruiter is guessing, hoping, frankly spraying and praying. That time is pretty valuable time that could be spent nurturing, engaging, educating candidates. Can you talk a little bit about the importance of not just the outreach but the actual nurturing process between finding candidates that could be a good fit and actually getting them in the front door?

Amit: That’s the other thing that it’s a very significant investment on the candidate side to prove to us that they are worth spending time on. The actual human part of the recruiting is also very, very intensive. I often find my engineering teams burnt out, with the number of interviews that they have to give to a target. Sometimes, we talk about why aren’t we programming to filter down the people that we interview? Then that’s asking somebody to invest a chunk of their time when they don’t even know whether they are interested in the opportunity or not. It’s just somebody needs to spend the time painting the vision for them and telling them why it’s going to be a great move for their career. That whole process requires a lot of time and compassion and passion. It’s great to know to narrow down where we spend that.

Mihir: There’s always a balance between do I keep sourcing for top of funnel, or do I spend time trying to help candidates learn more about the organization? The human element associated with that that you described is paramount. The tension between spending time doing one versus the other is a natural tension. I think it’s a place where AI can help really turbocharge to help recruiters function and help them allocate their time very differently rather than 50/50 or 80/20 being 20/80, along those lines. As you think about that and as we continue to think about how progress is made, it’s going out and finding new candidates and new candidates and new candidates. It’s just nonstop.

Then, there’s this massive repository of candidates that have been found that maybe were lost or came through the door at the wrong time. Now because of the growth of an organization, they could be a great fit for that role. How do you currently go back to that repository of people who’ve either indicated interest in your organization that maybe weren’t the right fit at the right time or didn’t respond because they didn’t know who you were? Now you’re at a place where people like, “We should re-engage.”

Amit: I think, sadly, the current state is that unless somebody just remembers to do that, it just doesn’t happen. There is a lot of lost potential in there that we haven’t tapped into. Every once in a while, it’s either one of us remembering that “We talked to that person. It didn’t work out that time, but it could work out now.” Once in a while, we’ll go and look at older feedback and see where whatever there is potential for the growth for you or for them. Because of the management growth, they could be a good fit there. Right now, it’s very grave, human-centric and walking on people’s mindshares.

Mihir: Do you have the time to go back and think about previous candidates who you interviewed 12 months ago or 24 months ago? I mean what you’re describing is it’s a highly manual process. I ask that because not only do you not have time to do it or bandwidth or the mental resources, but the current process is not set up to help those candidates learn more about the organization at the right time.

When you think about identifying internal candidates on your team and as you worked at places like – It’s very different when you’re at a 10, 20, 100 even 500-person company. When you start to get to 1,000, 2,000, 5,000-person company and you saw this at Bing – you saw this at Google – how do you identify people across the organization that could be great, internal fits on your team and then also provide opportunities for people on your team to find other places within the larger organization for mobility? Can you talk a little bit about that?

Amit: I think where we saw growth at Google and Microsoft was a lot of willingness organization’s part to allow this kind of mobility. The process of matching like people wanting to move and people wanting to hire was no different than – maybe a little bit more information available but then without two, you don’t have the ability to go look at everybody’s profile and things like that. The process was no different than external recruiting where people who advertise their job descriptions and hope that somebody applies would mostly reach out to teens where they had friends. They had used that –

Mihir: As you describe that, it sounds like that’s what you said earlier. You’re almost limited by the networks that you’ve already created or this. It’s difficult to look beyond those. As we see this accelerating plan of people migrating jobs within from ten years to five years to 24 months or 19 months, retention becomes incredibly important. As you continue to grow ThoughtSpot, how do you think about retention versus recruiting, the inflow on the spigot versus stopping the bleed on the other end? Can you talk about your approach to retention versus recruiting?

Amit: I think that most people like to do two things, ‘A’, that they’re making an impact and they’re growing while making an impact and, ‘B’, that they’re being treated fairly. Having a great team, people being treated fairly is kind of a no-brainer. You have to do everything you can to make sure that they are being treated fairly. Then, the other piece is around just giving people the right opportunities to grow and learn and things like that. That’s where internal mobility becomes one important factor.

There’s some people you just love to keep pointing their skills in one direction and keep working, but there’s some people who love to learn new things and try different things out and just go in and figure out what works for them. For that, we are – I feel like there’s a little bit of tension between you don’t want somebody switching teams every three months or six months. When they have reached a certain level of maturity, we are very open to internal mobility.

Mihir: At our webinar yesterday, we spoke with Ashish from Tata Communications. The way he described their approach is external recruiters have no problem reaching out to our staff trying to poach them. Why shouldn’t we also be able to use our own talent as a repository to find growth opportunities? Tata’s at a very different scale and stage than ThoughtSpot is. How do you, right now at your organization and as you think about the next three years or five years, proactively identify the people that you want to move to different parts of the organization to give those different experiences versus reactively having someone come to you and say, “I really want the –

Amit: What we’ve done is we’ve been very open about the staff that once you’ve spent a year, year and a half, we will be very open to moving. Once you’ve reached two, two and a half years, we actively want you to think about whether you want to stay on this team or you want to try something new and learn something new. It’s okay if you want to stay. It’s okay if you want to move. If you haven’t thought about it, we would encourage managers to bring up those things, whether they are at the best place that they could be.

Mihir: It’s a lot of burden not only to manage someone’s existing workflow and priorities but then also think about the best growth on their behalf especially for junior staff that may or may not be adapt at proactively asking for new types of opportunities. I want to pause here for a minute. I want to turn it over to Jason for a demo. We’ve had a couple of great questions come in through the webinar. We’d love to dive into that after the demo. Hold on for just a minute, and then we’ll dive back in. I’d like to introduce, at this point, Jason Gray. He’ll be giving us a demo of the Eightfold platform. We’d love to dive into it.

Jason: Great. Thank you for very much for the time today. I should be sharing now. As we go into the demo, the key takeaway here is a deep AI matching and how you see, throughout the platform, it relates to the candidate experience, sourcing and screening, the candidate nurturing and internal employee mobility. Underneath all this is smoothing out bias and creating more diversity as it relates to pulling in – sourcing talent, nurturing talent and so forth.

When we look at careers pages and you can look at any career page, they’re very confusing and are keyword-searched. You search for ‘software engineer’, perhaps, and you can find hundreds and hundreds of openings at any particular company you choose. Here at Eightfold, we really to improve that candidate experience. We’ll see that here when I simply go click ‘apply’, it brings me to a window where I can upload my resume. Eightfold’s going to automatically match my applications of roles that will best fit my skills or experience. This is on the actual Eightfold site. Our customers use this. I go in, and I grab my resume. We’ll do a software engineer. I upload this. We’re automatically seeing the deep matchings as we’re parsing the resume, applying it to all the different job descriptions and how they’re calibrated.

We’re predicting, in this case, Amy Jones, predicting what she’s going to do next. This is the same experience that we see when we do referrals, employees do internally or when an employee goes into internal job postings we’ll see later in the demo. I can see full-stack engineer, product management leads, senior front engineer so two software positions that I’m interested in. I actually have some product management experience in my previous engineering roles. Obviously, it pulled that in not obvious to me, but that’s why they matched it. These are the best positions out of all there at Eightfold, so I’m going to apply to both of them.

Then, as part of that candidate experience, we can also ask questions. I can state them specifically or even have text-based questions. I can go ahead and submit my application. Then, we can continue to enrich that with blog posts that relates to their position and other individuals they might know. That makes the candidate experience better. They get email notifications and updates through the Eightfold platform that our customers are leveraging.

Now, as it relates to the sourcing and screening side of the house where we deal with recruiters and hiring managers, what do they see? We really want to promote ease-of-use so that they’re not spending all this time going through physical resumes, going onto job boards trying to find individuals. Rather, they come in right in the morning and then get right to work with the best matches available. As we land here, we see all the positions on my right-hand side that I’m managing.

Then, I see a newsfeed of updates on the various positions. For software engineer, here I see there are 17 recent applicants. I see four leads that stayed in the pipeline where is the top 30 percent senior software engineer at Airbnb. I also see 26 new leads and two likely to respond.

We actually see some of that AI matching happening here because we’re looking at the career trajectory, the skills, the titles to match the right open positions. We also look at some of them have been in that position for a long time, thus compared to the 10 million resumes that we put print against that says there are two that are definitely looking that they most likely would move. That’s interesting to me.

Now, as a recruiter, the first position I want to work on here is the software engineer because I’m excited as I have a whole host of new applicants. I can see new applicants, 167. I also have leads here, 174, 176 now. It’s building now constantly because it’s pulling from the applicant tracking system. Those are people that might have applied in the past.

First, before I go to the people that applied in the past, I want to look at the people that are fresh and new and interested. I can quickly, if in a scenario where we’re working on diversity where we have, in this case, a very skewed team of men, perhaps I want to look specifically at women. I can look for those women. I can see that if they’re a top US school, top Canadian school. If I mouse over, it will also show here the relevance. As we’re doing this matching, we’re matching in skills and titles and ideal candidates either ones that have already worked here at my company or perhaps that I’ve pulled right down from a job board that this person would be perfect. The school elements work and so forth and so on.

Now I’m going to uncheck diversity ‘women’. I’m going to see just the top- ranking here. I see Ankur Garg. That’s a new one. I’m going to go ahead and drill in and work down my list and see what is Ankur like? He’s a lead software engineer. He’s got eight years of total experience. Then, we see some more of the matching eight years of relevant experience to this position. Everything he’s done is relevant to this position.

Then I look over here to the highlights, and this is where really interesting things start to bubble up. He’s the top 30 percent senior software engineer at SnapView and top 40 percent lead as well through his different positions he’s held there. It’s shown that what does that career growth mean? Well, it took him 6.9 years to get there compared to his peers that took 7.9. If I can see these top percentiles, that can help me hone in further on an ideal candidate. A lot of times, these candidates also might be – maybe he just got his Ph.D. who maybe had a year or two experience before that after undergrad, and he’s also had some internships. He’s really buffed out his resume.

Eightfold’s going to be able to identify that because we also bring in social elements like GitHub. Do they have a lot of followers? Are they committing or doing repos with projects and Stack Overflow? Are they asking or answering questions, and what are they answering questions about? That, I see right here those are major components of what we care about for this particular job opening. As I scroll down, I can see personal info, recruiting activities and emails sent and openings, any notes and also other experience they have. That experience they have, we’ll see these blue boxes. These are actual semantic pulls right from their resume that says they know MongoDB. They know SOAP. They know sales. They know finance, whatever might be that specific skill we care about. This deep AI uncovers these hidden meanings. I, as a recruiter, don’t have to go through resume after resume. I just get these top exact matches that are best for me.

Now I might want to get a bunch of these resumes over to the hiring manager, but we want to promote diversity so we’d like to mask those. I actually can make that happen through sending them over. Let’s just take a look at an example of Ankur here who’d be masked for the hiring manager. Now, I can’t see his actual name, not by suggest ethnicity. I don’t see, perhaps, the schools he went to. Any of this can be masked so that we can make sure it’s based on meritocracy of his experience or her experience in the application process. The hiring manager checks a couple of boxes on the people they like, sends it back to the recruiter who will then continue the screening exercise.

All of this would have taken a tremendous amount of time if I didn’t have this deeply embedded AI algorithms taking care of it. In my morning, I work the new applicants. I work the leads that are actually past applicants that Eightfold has also matched to this. I have my pipeline, people I’ve added there and also those ones I’m actively interviewing. Great. I’ve done some phone screens. I’ve scheduled some face-to-face interviews much more efficient than I have in the past. Now, I’ve got a charter. We have a lot of open positions that will be upcoming in machine learning. Wouldn’t it be great if I could go and send a notice out to those applicants in my applicant tracking system that are passive that I’d like to try to bubble up? We can do that by clicking on ‘new campaign’ here. I select, in this case, I’m going to share a blog or webpage but it could be alumni. It could be a specific opening or location.

This is literally like marking on, embedded right into the system. I’m going to go ahead and grab this great blog post from our site or from another area on the web that suggests our thought-leadership and would target the right people. All of the sudden, you’ll notice here that my target audience went down to just under 11,000. I take that back out again and press return, I would see that number further increase back up. I might want to say, “This is a diversity scenario. We want to target women.” It went from 10,000 to 1,800 now.

Then furthermore, I want to just select a degree like a master’s degree. Now, we’re down to 782 possible passive applicants in our ATS. This is all about taking your existing investments, all of that money you’ve poured in overtime trying to recruit people and hire people and thus bubble those up and engage them for new campaigns. The beauty behind this and what our customers are seeing and you’re going to see here when I drill into a particular campaign that’s already run, I see the number of sent campaigns. I see the opens.

On average, our customers are seeing over 50 percent open rates because of this deep matching of how we parse the content of that blog post that I showed and matched it to the right people in my applicant tracking system. I can further see what that email looked like, the audience deliverability of that email and ultimately the recipients. I can see how engaged they are like Nicole here, two opens and 20 clicks and also Nivien and Michael.

I’m going to go ahead and share these right now with the hiring manager and let them know that these look like awesome applicants from previous attempts to get them on board, so let’s go after them again. Now, all of a sudden, the ability to nurture candidates and do it quickly is not a dream. It’s an actuality.

Finally, let’s go take a look at the internal mobility where I am in the sales department. I’m looking are there any positions that are relevant to me? Here we go. It has predicted that I’m ready to move for up to a director of sales role. I can drill into that, look at those particular jobs or job or even look at other jobs that might be of interest to me, my applications, my referrals so I can take a resume, upload it. There may be hundreds of openings. It doesn’t matter to me because Eightfold will automatically match them for me. Career planner, I can find mentors within my company as well as projects that might be of interest to me that I would want to improve my skills and so forth.

Now, finally, I also might be an administrator in the HR department. I can come in here and look at people in our organization and come down here and say, “Who is a high attrition risk,” people that have perhaps been in positions too long. Thus, we want to make sure that we retain this talent and keep moving them forward. I can look for those and be proactive. Thanks to the matching identify their career trajectory as being an opportunity for us to improve. As we look at this, the candidate experience, the sourcing, and screening, this wonderful nurturing and employee mobility has this deep AI matching to drive tremendous efficiencies and ease-of-use for all those involved in the process. I’ll kick it back over to Mihir to finish up.

Mihir: Jason, thanks so much. I think you touched on some key elements that Amit covered during the earlier part of the presentation in terms of challenges and opportunities. I’d love to kick it back over to Amit with a couple of questions from the audience. As we talk about AI and just broadly as an industry taking over key functions or key parts of our daily life, a question that came is will AI replace humans? Then he says broader than necessarily just recruiting, how do you think about AI versus a person ally?

Amit: I think it’s just a silly thing that gets a lot of interest. I don’t think we are anywhere close to AI being anything but an extremely powerful tool to help everybody realize their further potential. As this society moves through more technology, there’s always some jobs that get replaced with some other jobs. Throughout this, some people feel the pain. It happens with automobiles. It happens with an industrialization. It will happen with AI as well. I don’t think that we are moving towards a dystopian world or anything like that.

For example, my company at ThoughtSpot, what we do is that we make – we enable anybody to be able to integrate their data. This used to be not possible before for most of recruiters. They relied heavily on Atlas or people who – so there’s all those questions of what’s going to happen to my job as we deploy ThoughtSpot? Time and again, what we have seen is that these people actually get promoted when ThoughtSpot is deployed. Instead of doing the tactless job of repeatedly doing approach, then they’re doing higher-value things. They drive more value for business. That’s a base for them to get promoted to higher responsibilities.

I think similar things is going to happen with recruiters as well. They will just make a recruiter a lot more valuable to the atomization and a lot more effective. They will be able to source more candidates, choose more candidates and be better at the job.

Mihir: The human element you talked about earlier and AI is a tool to help recruiters spend more time with the right candidates, that is actually probably a way to hone that in a little bit. Tools like this should help them scale. When you think about how you direct your recruiting team through the growth and you talked about diversity earlier now, another question that rolled in was how do you think about AI for diversity? Does it promote bias, or does it actually help with diversity? Can you talk a little about AI in the sense of –

Amit: AI is like a power tool. You can create beautiful things with power tools, and you can create a destructive course in power tools. You have to know your tools, and you have to direct it in the right direction. There is a case to be made that all that if you’re feeding to your AI engine is people look a certain way or come from certain colleges, the AI will learn the same thing and reduce diversity. They will be progressing that. It doesn’t have to be a reason to stop using the power tools.

Mihir: It’s interesting, as you said, I think that analogy is particularly cogent. A tool is only as good as the operator and then the direction that you’re pointing it. When you think about recruiting and the next phase of growth for ThoughtSpot, how do you communicate some of those changes of what you’re hoping for as you continue to grow from a recruiting perspective?

Amit: Right now, I think I’m not seeing anything else change other than scaling in recruiting. With scale, you have to change your operation. You’ve got to change the structure in which you recruit. The kinds of people you will recruit, the quality of people doesn’t change. The other thing that changes, at this point, we have enough named recognition and brand recognition that more and more people want to come and apply. There’s a lot more candidates to go through. There’s a much larger pool, and so you need to be more efficient with that. Other than that, I’m not seeing anything change.

Mihir: As you talked about scaling operations, you either scale with people or you help people with better tools. As we come to the end of our time here, I think you’ve touched on and have frankly gone deep on a few things that are core challenges for recruiters across the spectrum be it sourcing, screening through the funnel and then internal mobility use. You talked about the entire spectrum. I found this to be an incredible learning opportunity to understand how you go from an early-stage company that is fighting, scratching and clawing to get the next piece of talent to how do you now harvest a lot of that coming in. I’m looking forward to having future conversations to track how you’re managing and scaling that growth. Hopefully, it’s not a linear scale with people that process as you go from a few hundred to a few thousand to maybe a few hundred thousand in the future.

Thank you so much for your time. I know it’s valuable. We found it incredibly enlightening. Audience, I want to thank you so much for your participation today. This is the second part of a three-part webinar series. Please, tune in tomorrow at 10:00 AM. We’ll cover the third piece of content. I’m excited to continue this series. Amit, thank you, again, so much for your time.

Amit: Thank you. It’s been fun talking.