AI in Hiring: From Microsoft to Humanly's Recruitment Solutions

Prem Kumar, CEO and co-founder of Humanly, shares how the company helps hiring teams screen, schedule, and interview candidates efficiently and equitably. Focusing on early hiring stages, Humanly aims to address biases and improve the candidate experience. Kumar shares insights from building the company, highlighting the importance of responsible AI and the right features. He discusses retention, employee engagement, and challenges in high-volume recruiting, offering advice to aspiring entrepreneurs to overcome fears and take action.

Takeaways:
🤖 Candidate Engagement: Candidates appreciate the opportunity to engage with a chatbot rather than being ignored.
⚖️ Fair Screening: Humanly focuses on providing a standardized and efficient screening process, ensuring equity and fairness.
🌐 Ethical AI: The company aims to be a leader in responsible and ethical AI in the recruiting space.
🌈 Inclusivity: Reducing bias in the hiring process can create a more inclusive workforce.
🏅 Team Culture: Mentorship and building a strong team culture are crucial for success.

Resources:
Podcast Feedback Form
Podcast Guest Nomination Form
Humanly
Connect with Prem: LinkedIn, @PremKumarTweets

About Prem:
Prem Kumar is currently the CEO and co-founder of Humanly.io, a venture-backed generative AI platform that empowers hiring teams to have more effective and equitable job candidate conversations. The process should be driven by humans while automating only essential tasks. He has previously led product management and design teams at TINYpulse, an employee engagement company focused on real-time people data for building world-class cultures in organizations. Before his time at TINYpulse, Prem spent 10 years at Microsoft working in various product capacities including HR Technology, New Ventures, Dynamics 365, and Office 365. In recognition of his work, he has received several honors such as Geekwire's "Startup CEO of the year - 2023", being named a PSBJ 40 under 40 honoree, Forbes NEXT 1000 honoree, TAtech Top 100 leader, Top-100 HR Influencer, and was listed on "The Most Inclusive HR Influencer Li

Chapters

• 0:00 Introduction and Overview
• 3:49 Humanly's Role in High Volume Hiring
• 4:57 Working with Mid-Market Companies
• 6:37 Categorizing Candidates and Providing Outcomes
• 8:33 Improving Candidate Experience and Equity
• 11:43 Building Humanly and Surprises Along the Way
• 12:50 Addressing Inequities in Interviews
• 14:27 Building the Right Features and Product Market Fit
• 17:59 Humanly's Series A Funding and Growth
• 21:23 Using AI Technology for Effective Conversations
• 22:15 Guardrails for Responsible and Ethical AI
• 27:04 Career Path and Early Experiences
• 27:30 Interest in Technology and HR
• 31:06 Informatics Degree, Interdisciplinary Opportunities, Practical Experience
• 33:09 Background and Early Career at Microsoft
• 33:44 "ThinkWeek" and Prem's Transition to HR Tech
• 35:37 The Importance of Retention and Churn, and Reverse Mentorship
• 41:08 Leaving Microsoft and Starting Humanly
• 42:39 Startup Experience at Tiny Pulse
• 46:18 Building the Humanly Team
• 50:16 Standardized Screening for Fair Hiring
• 52:45 Team Culture and Hiring Beyond Technical Skills
• 54:52 Future Goals and Value-Aligned Customers
• 56:20 Prem's Advice for Aspiring Entrepreneurs

Transcript

0:00

Hi, my name is Joan Palmiter Bajorek. I'm on a mission to decrease fluffy hype and talk about the people actually building in AI. Anyone can build in AI, including you. Whether you're terrified or excited, there's been no better time than today to dive in. Now is the time to be curious and future -proof your career, and ultimately, your income. This podcast isn't about white dudes patting themselves on the back. This is about you and me. and all the paths into cool projects around the world. So what's next on Your AI Roadmap? Let's figure it out together. You ready? This is Your AI Roadmap, the podcast. Hey folks, this is Joan saying hello. A little intro to the episode. So in this episode, we hear from Prem Kumar, who is the CEO of Humanly. if you're actually watching this episode on YouTube, I actually am in their building. There's a background of different colors and you'll notice hey, I'm not in my regular office, because I went and visited their office. Anyway. So in this episode, I speak to Prem and hear about his experience of really figuring out where he wanted to expand his career and the problems and opportunities he sees in the recruiting space. And if you've ever looked for a job, it can be so infuriating inputting your data into this box of a black hole and you're like, where does all this go? And he really talks about how it's potentially frustrating for the... businesses who are looking for the right people to hire. And so thinking of making this more equitable, more seamless, how people can be happier even when they get rejected. A polite rejection is sweet, right? You're like, hey, there's so many awesome applicants. We super appreciate you. Talk to you again or never. But like, I think being respected in this time. I mean, the company's named Humanly, right? We're literally talking about a company named Humanly. So I'm really excited for you to hear this episode. Did I mention he also went through YC? The most famous accelerator in the world. Anyway, I think you're gonna really enjoy this episode with a very humble and cool founder. and one more thing before we dive into today's episode. Can you believe season one is almost over? we're so excited to announce that we've started recording season two. We have some amazing guests to share with you. And as a data -driven team, we need your input to make it even better. Data, data, data. What did you love about season one? What could be better about season two? Are there guests you are desperate to hear about? Companies you'd love us to feature? What questions do you have for us about AI, careers, entrepreneurship? Do you have any feedback to share? We would love your constructive feedback. So let's make season two epic together. Visit the form in our show notes at yourairoadmap .com / podcastfeedback Let us know. This is our first podcast. We are learning and growing with you together. Your questions and feedback might be given a shout out in a future episode. I would love to be calling your name. just again, one more time, that is yourairoadmap .com / podcastfeedback. Okay, thank you. Let's dive in. Hello. Hi Joan, how are you? I'm well, how are you? Doing well. could we start with, could you introduce yourself please? Yeah, my name is Prem Kumar. I'm CEO and co -founder of Humanly. We help high volume hiring teams screen schedule and interview more effectively and equitably. Ooh, I dig it. Okay. could you unpack that for us? Tell us more about Humanly. Yeah, absolutely. So we've been... around since about 2019. I've been in the kind of AI and broader space for many years, about 10 years or so. Humanly specifically, we're focused on kind of the earlier stages of the hiring process. So we'll work with companies that are getting high applicant volume. These are usually entry -level roles, maybe mid -level roles. We're getting thousands and thousands of candidates. So we help hiring teams screen. that high volume of candidates, we help them schedule them. And then we have tools that will sit in on remote interviews via Zoom and other tools to help measure how we're showing up. But really, anytime you're having a two -way conversation with a job candidate, we want to make it more efficient and more equitable. Cool. Well, and so if I'm thinking about, I'm not a recruiter, I have no background in this experience, but I certainly have interviewed at companies. I thousands of candidates, these have to be relatively big companies. to be having that many inbound candidates, right? So are we talking about like the Googles and Amazons of the world? I don't know if you can disclose your customers, but. Yeah, happy to give you an example too. Generally speaking, we're targeting companies with about 1000 to 5000 employees. mid -market, we also do work with some enterprises. So five to 10 ,000 or 10 ,000 and above from an employee standpoint, a little bit of a different part of our product they're using, but. One example on the mid market side is a Seattle company Moss Adams. So in the past they were getting about 4 ,000 university candidates that apply through university recruiting in a two month period. And they just had the human time to engage with a much smaller percent of those candidates that applied. So we allow you to have a conversation through our chat bot with everyone that applies within 24 hours, screen them and then schedule the ones that are most. the best fit for the role. That's awesome. What? 4 ,000 candidates in two months. Honestly, I feel overwhelmed and want to run away from that. How does the chatbot filter? What does it look for? how that go? Yeah, yeah. So generally, yeah, it works. some of our customers have many more than 4 ,000 in two months. So it's definitely a... High volume game. So I think it's a couple things. I think it's also about the candidate filtering the company. I feel that, you know, if hiring teams had unlimited time, money and resources, every initial interaction with a candidate would start with a two way conversation. So we use the chat bot to answer candidate questions as well as screen and filter them. So then when they get to the next stage, they already know, hey, I've gotten my questions answered. This where I wanna be so there's less drop off later in process. But to answer your question specifically, yeah, a lot of this depends on the company. We have question banks and what we're basing this on. Sometimes it's the most basic of items that we're looking for to just make sure there's a fit. So sometimes candidates might not know, hey, here's how the work -life balance is set up or here's. the requirements from an H -1B visa standpoint. So they're applying anyway. So we'll start with kind of the general knockout type criteria. And then we'll go to more of a question set, depending on the job type, it might be more behavioral questions, it might be questions that have to do with skill, things like coachability. But that's kind of how the process is set up. And we believe in an outcome for every candidate. So it's not just about a fit now, but there might be a candidate We might put in a, hey, they're a silver medalist, so they are qualified, interested, and available for the most part, but maybe they're missing one criteria. Do we get re -engaged with them later? So it's about talking to everyone and providing an outcome, whether that's an actual next step in the interview process or a future opportunity. Silver medal candidate. like, right. that a ranking system or is there a dashboard that people... consume on one side or like, how does that look? Yeah, so there's three categories. So we'll either say, hey, this person, Joan meets the criteria we're looking for, we want to want to schedule them. So the bot would then schedule them directly on a recruiter's calendar. The second is more of that silver medalist. So they're a good fit. They seem to be generally meeting the criteria, but maybe in the university hiring case, they aren't getting their degree for a year. So we need to keep them engaged and wait. And then there's one that maybe it's not a good fit for other reasons. Maybe they don't have basic qualifications or requirements such as visa or travel requirements or things like that. So we'll put them into three categories and we'll put them into the applicant tracking system. So what we're doing is now instead of looking through 4,000 You're now meeting with the ones that are populating on your calendar. And from a candidate standpoint, instead of in many cases, just not hearing back at all, everyone is jumping into a process and talking to the company via our technology within 24 hours or so. Sounds like a very adult thing. Or like we have a system. There is a filtering, almost like there's more equity. almost with the candidates or I feel for people who are job hunting applying to 200 opportunities. And like you said, almost not hearing back from a lot of them, regardless of how talented you are, frankly. Can you tell us more about like, you know, how you've been building this or how you even started with this problem, I guess, might be a place. Yeah, for sure. And it goes back to the E word that you used around equity. So when I was graduating from University of Washington, we had a lot of folks coming from hiring companies to campus to interview. I majored in informatics at the time. There was a lot of tech jobs. And it was right before, it was in 2006. right before the financial situation that happened shortly after. But it was a hot market for job candidates. what I would see is I'd apply to a lot of jobs and generally not hear back at all. then the ones that I did get interviews with, they'd come on campus and my colleague and I we'd go and talk to the same panel. We had the same experience, the same degree, and we just get asked completely different questions. They were grilling her on her technical skills that wasn't coming up in my interviews. It just felt very disjointed, so very hard to get to a human. then once you got to a human, it seemed like there was no method to the madness. as I got a job at Microsoft and being in HR tech and seeing the recruiting teams and hiring teams, I realized it wasn't because these recruiters and hiring teams were trying to be biased or just not wanting to talk to everyone or engage with everyone. They just didn't have the tools to engage at scale like salespeople might have when they're dealing with prospects and things like that. That's so interesting, right? Because salespeople are enabled with, they're wigged out. They're teched out, but kind of the recruiting HR side doesn't have these tools, is what you're saying, and kind of cleanly filled that. Yeah, no, if I were to tell our marketer, hey, drive a million eyeballs to our website, but sorry, we just have enough salespeople to demo, you know, -tenth or one -hundredth of the people that want to buy our product. That's what's happening in recruiting. You have people raising their hand. that are interested in your employer brand. They could be a good candidate now, they might be a customer. So a lot of our customers are B2C and candidates oftentimes are their best customers. You take Disney, for example, not a customer of ours, the average Disney job candidate spends eight times as much money at Disney parks as the average consumer. So. treating candidates well is really good for bottom line. It's good for your business. It's good for your employer brand as well. Absolutely. Yeah. I didn't think about it that way, but right. I was gonna say your target user, your employee and also your consumer. That's fascinating. Well, before I ask you kind of more your career and back, I guess to Microsoft days maybe, as you've been building Humanly, what are some surprises along the way as you've been building it out? We'd love to hear more about that journey? Yeah, when we started. So the tool again, there's a chat bot that screens and schedules and then afterwards it joins a Zoom meeting and it gives feedback to recruiters. It takes notes for them. So it's really an assistant that does pre -screening and then helps along in the interview. Some of the surprises, One of them was when we initially pitched our idea, there was a lot of feedback around do candidates actually want to talk to a chat bot? Would they? rather talk to a human and that's obviously a very common question. What we realized in high volume is it wasn't the difference between talking to a human and talking to a chatbot. was from a candidate standpoint, it was a difference between being ignored and engaging with the company. So what we found that surprised me is candidates that don't get the job will rate their experience about 4 .7 out of five. So what we found, even people that were told they aren't a fit, love being able to jump in early, learn about the company, get screened. So that was one surprise. another surprise, this is more experimental in the technology we're building, but we did a lot of analytics on Zoom calls, on Teams calls, and just the level of inequity was very surprising. with one of our customers, We found that junior female candidates were getting eight minutes less to talk in an interview than their male counterparts. So we're able to measure, the interviewers showing up on time? If there's a pause in conversation, are they jumping in and interrupting? And we found in certain cases, people of color and women were getting unfair interviews. So that's more on kind of the experimental piece of what we're doing in Zoom calls. But yeah, those are two things that... You know, you read about the inequities, but seeing them firsthand with data was a surprise. That's fascinating. Well, it seems like it parallels a bit your own experience of hearing your female friend and how she was being interviewed versus actually having data potentially on at scale or growing being interrupted or otherwise. And is that eight minutes less in a 30 minute or Like what percent Yeah. So, it was across both in this research, it was across both 30 minutes and some of them were 45 minute calls. It was about half an hour. Okay. Yeah. That's a pretty big chunk. Yeah. I really don't believe that fundamentally humans are like trying to be biased or trying to be rude to different people. think we unfortunately learn these things in our culture and societies. You know, as you see these differences, are there pop -ups, there like intervention tools or is it right now at a stage of collecting data, learning more? it, hey, do you to hand them the mic again? I'm wondering how you think about different feature build -outs. Yeah, I mean, I think understanding our role and technology's role in a human -driven process is important. And I think it will continue to be human -driven. I do not think technology will solve all these problems. What happens once you're off the Zoom call and you're working with these people. I think there's there's areas that we can give insight and course correction. The way our tool specifically works is right now we're very much focused on the front side of this through the chat bot. So we can ensure things are happening in a standardized way. have our AI through the chatbot conversations. We have our AI audited by a third party. We're careful in the words that our chatbot use to make sure we're not using gender language or ageist language. So we can control a lot of it in the first touch, which is where a lot of this bias happens in high volume. Because if you're a recruiting coordinator and you're maybe speaking to about 10 to 20 % of the people that applied, Even those folks that you're talking to, have about seven calls a day. It's pretty, you don't have a lot of time to really, there's a lot of opportunity to make mistakes. And I think by automating the first piece, we can guarantee a level of standardization. And then by having measures, the Zoom piece I was talking about was more so done via measurement. We're not providing insights like inside of the call. Although some companies are doing that and I have some thoughts on that. yeah, generally speaking, think right now it's about correcting the problem at the beginning by making sure everyone gets a chance to chat with the company and then providing some data back in the next step. Yeah. Well, it also, it sounds like even there's kind of a reputation or like this company actually gave me a chance to like at least say my thing, like even if they're going to be screened out or as you mentioned, they have a higher appreciation of the company. sounds I've worked in chatbots a long time, so it's interesting to hear this type of use case. Are there other learnings along the way that you've had as you're building out different features or things people really wanted that you weren't expecting or otherwise? Yeah, I think as it relates to product, think the best decisions we've made have to do with We have chosen not to build versus what we've chosen to build and you're obviously limited as a startup with resources and the saying goes at startups, so die of starvation, they die of drowning in ideas, not die because they don't have enough ideas. So I think early on choosing what not to build and we did see, obviously we couldn't have predicted how quickly. some of the generative AI movement would have happened and what's happening in the LLM space. But we did know that certain parts of chatbots, for example, the infrastructure around making technology good at talking to someone, we felt that was becoming commoditized and things like sentiment models. And there's a lot of pieces of machine learning and AI that we felt were going to be commoditized. They happened quicker than we thought. So we figured, like, how can we focus on proprietary data around a good recruiting conversation and use that specifically. So I think that was one thing that was important to us, just knowing not to go too deep on the stuff that we felt could be solved by bigger players or other players and really focusing on our lane specifically. Absolutely. I agree with you. Ideas, ideas, ideas. What are you actually going to attack? What are the next steps? Is I think, a thing for most startups. And what series is Humanly these days? Yes, we raised our Series A about seven months ago now. yeah, the companies were about 40 folks. We grow in pretty quickly. We're working with over 300 companies and making their hiring process better. Wow, congratulations. Series A is no small feat. So that's amazing. When you said you started 2019 and we're recording this in 2024. So it's been, do you feel like you knew kind of what you thought you'd be building? Does that match today? Yeah. So like, yeah, we incorporated in 2019 and we really started like, and at that time I was, you know, we weren't full time yet. I think that when we really started selling and building was more so mid 2020 towards end of 2020. And at that time, What I would say is, to be honest, it kind of goes back to the other focus thing where over time building not something different, but more of a subset of what we wanted to do at the beginning. And it's actually the vision is pretty similar, but where we've focused, it's been more about just building, being really good at a smaller set of things than a bigger set of things. that process over time is kind of one of focus has been important for us. Absolutely. Well, and you said 300 customers right now or like, and if 300 is a good set of people saying this is what we want or like willing to buy. Yeah. Yeah. So we, yeah. So we have 300 companies using us either through direct sales or through partners. So we have, yeah, happy to name some of them as well. But yeah, certainly we've gotten product market fit is to me like a, it's not necessarily like a a destination you get to and you never have to worry about it's more of like a process and an ongoing way of building. So I would never say we have we've like found product market fit now it's over and we figured out everything. But yes, I do think we found something that a lot of companies are engaging with. We have I think five to 7 ,000 candidates a day that are having more equitable and efficient interviews and screening conversations via Humanly now. that's been exciting to see that impact. Yeah, that's amazing. Do you want to mention a few of your customers, ones that people might know? Yeah, for sure. Yeah, so we're all over the US working with companies like Taco Bell, World Bank, Krispy Kreme. We, Moss Adams is one here in Seattle and accounting for a big accounting firm we worked with. Generally speaking, we have kind of two segments of customers. One is the quick serve restaurant, QSR space, as well as like deskless workers in general. And then the other piece is professional high volume, which is where a lot of our more recent features are. So these are banks, insurance companies, accounting firms. Interesting. Yeah. When you started listing off the companies, I was like, whoa, is this tons of CPG and food? Like I wouldn't have been my first guess actually. So that's cool. But they certainly need this as well. Well, okay. You mentioned kind of the niching down, but as you think about, you know, the next five, 10, kind of where do you see Humanly headed? Where do you see this part of the field headed? I'd love to hear. Yeah. Really what we want to do at the end of the day is anytime a candidate is having a two-way conversation with a company, we want to be there, whether it's automating it through a chatbot if it's an initial conversation or it's helping a human have a more effective conversation down funnel. So that is our goal. Anytime you're a candidate talking to a company in a two -way format, we want to make it more efficient and equitable. So that's been a big focus of ours. And then I think just from a technology standpoint, we very much are trying to build kind of a proprietary set of data and technology around making those conversations more effective. So how do we get a candidate to convert? How do we answer this question in a way that's going to be effective? How do we build guardrails? So if it's a candidate in Colorado where there's certain pay transparency laws, the chat bot is effectively answering that. And then I think the other piece is we want to be the leader in our space as it relates to responsible and ethical AI. And I'm happy to talk about that more, but clearly at this time, there's a lot going on. We want to make sure our customers have the right guardrails to take advantage of the technological changes, but do so in a way that works for recruiting. Yeah, that makes sense. Well, I think a lot of people, as you mentioned with generative AI, we want to implement it's going maybe faster than some assumed, but then also like, is it spitting garbage, know, like hallucinations. I've had so much fun being like, what? This isn't real. When you think about, and I love the example with, you know, Colorado has specific laws about pay transparency, California has their other like, it's, you know, United States is so fragmented sometimes when it comes to different laws. How are you thinking, maybe maybe just go with hallucinations for the minute, like, how do you all think about guardrails in that respect? Or do you have content editors? Like, how do you, how do you work with Yeah, so we do feel so we at the end of the day, we want to give our customers flexibility. So we will plug into a bunch of LLMs and there's only certain scenarios where there's where we need to go beyond kind of the curated set of the knowledge base that's internal to a company or our specific knowledge base. In those scenarios, we're really careful. We know that LLMs like OpenAI are trained on the Internet for most cases and the internet is fairly biased. ChatGPT is kind of a white male in many ways because the internet was generally created by English speaking white men in many ways. I think about 70 something percent of the pages on the internet are English and the second language is Russian at like 6 % or 5%. So the corpus of knowledge that humanity has put online, we feel is going to have some innate bias in it. So first, like how we put guardrails on is we don't think, and I'm not just speaking about Humanly, but in this recruiting tech space, we don't feel the answer is kind of a thin layer on top of a third party LLM. We do feel a lot of this is going to be based on internal knowledge bases and whatnot. So the way it works for us is a lot of our candidates too, as exciting and cool as it is type whatever question and get an answer. A lot of this is you're on mobile, you're trying to apply rules based, like button choices get you pretty far. So when we go into more open -ended scenarios, which oftentimes if you're on mobile, you might not wanna like type a bunch of stuff. You might wanna select a bunch of things, which is quicker and easier and more efficient for these entry level jobs. But when we do get in context that are more open -ended, I think we'll first see if there's a curated set of answers, basically either in our database that we've done with the customer or on their career page. And we're going to kind of pre -created sets of information. I think the other guardrail piece is giving customers a flexibility to say, hey, this is an onboarding. We want the bot to answer questions about benefits, about career, about culture, but not about pay, not about these sort of things. So I think giving customers the ability to choose upfront. And then using the LLMs, I don't mean to sound negative. There's a ton of amazing things they can help with. But I think using that in the right context instead of just keeping everything open-ended. I think buttoning it up is... our process for training our AI is kind of similar to how you would train a recruiter to answer the questions in the right way. And I'll last thing I'll say is often... encourage companies if they're hiring AI to do what humans might have done in terms of talking to candidates. They should almost ask the same questions as they would if they're hiring a candidate. What's the training? What credentials does AI have? Where to go to school in terms of the data sets? Is it a good culture add or is it detracting from your culture? those are some of the things we think about. It's fascinating. Yeah, well, I'll admit, I assumed you were calling an API to something else and putting a foil on it. So it's interesting that you're like, hey, these data sets won't match our use case or, you the goals of the different bots that are being built out. And I definitely, having built several many, many, many bots in my day, having a bunch of text sometimes is wildly unhelpful. Like you only need the yes or no, or these are really the three type of answers I can work with on this data set or having to, unfortunately, throw out a lot of material. And yeah, if you're on your phone, you don't want to necessarily write a novel necessarily. that makes sense. Well, are there other things you want to share about Humanly before we jump into your career path? No, I think you asked some great questions. So I got everything I want to share about Humanly. cool, so you mentioned UW, University of Washington. Do want to start even maybe earlier? Yeah, so I got exposed to the job market high school or junior high, so I started doing different application processes there that probably left some feeling around how these processes can be improved. So first job there when I was in high school, I worked at Albertsons for about four years. and really enjoyed the time there, but it was also just kind of clear how just hard it was to find the right job. as I, went to college, I saw how technology at its best can help things happen faster, more efficiently. But it can also make things that are bad happen faster and more efficiently. And then AI has shown that as well. So I think if technology is a catalyst, what I saw is first we need solid process inputs. And I looked at there, I saw opportunities in HR and recruiting where we can use technology to have the good stuff happen a lot more frequently or automate things that maybe humans didn't want to do or could be better served doing something else. So started seeing that through my degree and meeting with folks and doing an internship at Washington Mutual Bank, which doesn't exist anymore, but I was part of the technology solutions group when I was in college. So that was my first internship. And I quickly saw some applications for technology within the HR realm. So that was one of my first, I guess, moments around seeing the promise of tech. Okay. Yeah. One, it sounds like you really started being the end user is what I'm kind hearing of. Like, did you happen to apply to Taco Bell? have to ask you. Not at Taco Bell. Yeah, I applied at a lot of places. I was walking down the street at the University of Washington campus, just going to any like restaurant There's this pottery place I walked into, bookstores and going anywhere I could to try and find a job. it's funny just looking at my dad's story. This was more applying to college, but it was back at a time where everything was on paper. So he was in India and he was wanting to apply to universities in the US. It was before the internet was there. So he would go kind of door to door to see if anyone had friends of his or. people from his undergrad and see if anyone had extra paper applications and he found an extra one to USC. He filled it out and he got in and he moved here. very interesting to see just this whole application process and how much it's changed I felt then, I feel now that, again, with unlimited time, money and resources, any application process will start with a conversation. We just haven't had the... technology to do that. So we use shortcuts like resumes and things like that to kind of get a quick look at who someone is. But really we want to create a way for you to have that two way conversation out of the gate before you move forward. Yeah. And it's interesting. You also mentioned your dad or just like paper. Because my mom applied to her PhD on paper. Like there's so many things that I think about, you know, just a generation ago, like the digitization of things. and then you're taking it a step even further and like beyond digitization, how can we make it better? Like, well, I guess Humanly I'm like, this makes it more human. It's like conversational aspects. It's a good, name. and I will say, are there any full circle moments of interviews that you had previously that now you're like, you were the end user previously that maybe didn't have the best, you know, interviewing process and now you're actually working with that customer? Yeah, there's a few. I won't name any specifics from a negative experience standpoint, but definitely, you know, I've applied to jobs and there is one in particular, this is more retail and I was told I didn't have the right look for the job and they're actually a customer now. So hopefully we can help them reduce the bias in the early application process. very cool. Well, and or not very cool, excuse me. It's not cool that they're biased. It's really cool that you're building tech that helps folks. So you went to University of Washington. You were getting this informatics degree. Do you see parts of that that lead into your work now? Or was it more working at Microsoft that was the next step? Or how do you tell that story? Yeah, know. Good question. And 100 % started with the Informatics degree and also the folks that I met, whether it's professors or colleagues of mine. One of our advisors right now at Humanly actually was in the Informatics program ahead of me and he kind of helped me through and was a mentor for me. So I think that I really enjoyed Informatics because it was You had the technical element where I took a lot of classes in the CSE department, the computer science department, but and then there was also a little more human centered design project as well as a project management component. So I got to do a lot of like practical things and it's kind of run full circle or now I'm on the board of informatics has a program for students that are interested in starting a startup. So they have an entrepreneurship program. So I've gotten the chance to speak there and meet with folks that were similar to the phase I was at at that point. And it's just amazing to see the amount of resources folks have to get into the entrepreneurial ecosystem early. yeah, informatics was kind of the catalyst there and University of Washington has a lot of programming. So there was a business school startup competition. I wasn't in the business school, but went through that, didn't win it, but it was a great experience. so yeah, I definitely took advantage of what was there. And Microsoft is really when I got my first foray into HR tech and recruiting tech. Yeah, well, and we both just happened to be University of Washington grads. I certainly, love the opportunities of that campus because there's so many different schools and that even you're talking about like the interdisciplinary -ness that you can just walk over. and experience stuff at the business school. You can just walk over and check out computer science, which definitely speaks to my experience. Go UW. Great university. Yeah. and what were you working on at Microsoft? Yeah. So I started Microsoft in a customer -facing role as an account manager. So I'd worked with a lot of enterprise clients that were using Microsoft products at the enterprise scale. So that a good way of, there's a customer service element, a support element. So I got to really see how people were using SQL Server, Exchange, all those sort of things. As I evolved my career there, I really was interested in product and project management, those sort of things. But to me, the more important thing wasn't just like the job function, it was like the type of product I'd be working on or the type of project. And I actually took a ops role at Microsoft and I was actually working with our India team a lot of late hours and I started just writing down what I thought Microsoft could do better from an employee engagement standpoint, a lot of like HR topics from connecting with millennials. Now, of course, there's a lot of great stuff on Gen Z and at that time millennials were where Gen Z is now. So I started writing that down and then there was an internal process called ThinkWeek where you can submit your papers. So I submitted my paper about how Microsoft can better retain and engage with millennials. And the paper did pretty well and I got a chance to meet Lisa Brummel who ran our HR team at Microsoft at the time. she's now an advisor at Humanly. But I realized that there was maybe something to this passion project of mine. I then shifted into more of like an HR tech role and for about five years was running our HR portal at Microsoft from a technology standpoint. that was kind of the following, not just the title of the job, like product manager, project manager, more so that the thing I was interested in was people and data. So I found an opportunity to move into product and project type jobs within that realm. kind of where I met a lot of the folks I know now and really where I kind of started this HR tech, recruiting tech type of career path. Yeah, that's so cool. is, and maybe this is not how it goes, but what I'm hearing is you were working late at night, working your Microsoft job, documenting, this could be better. As a millennial, instead of a burn book of frustration, you actually turned that into something constructive and frankly extremely valuable. for Microsoft and they were open to even hearing about it. didn't know about their cool week for that. like, did you feel very disgruntled or was there some like kind of grungy aspect to this? I had a great team and all that. So most of the things you'd look for a good manager, good team, but I would say, yeah, there was definitely an element of not being completely fulfilled. And I would say if I wasn't doing that writing, I might have left, but I saw like I used writing down what I saw happening as kind of like an outlet and I didn't know exactly what I was gonna do with the writing at the time, but I started interviewing folks. I'd go to coffee with other millennials and take notes and I started seeing that, maybe there's a change we can make here in terms of how these, because Microsoft's also struggling with two -year retention rates of millennials and there was a lot of turnover and so it was a problem the company was trying to solve. And one of the things that came out of my paper was a reverse mentoring program that we had where folks from my generation at the time could mentor senior leaders. So I was mentoring, then I met someone else I ended up mentoring who was much more senior than me, but helping him manage the team of millennials that he had. eventually there was another program which I didn't have any involvement in, but I was excited by used reverse mentoring in a different capacity where it was actually junior female engineers mentoring male execs on how they can be better at engaging that population. So yeah, it was cool. It did start from maybe a place of being not extremely fulfilled and Microsoft provided kind of an opportunity for me to use that energy in a positive way. Absolutely. No, it's super cool. I love I've heard about this reverse mentorship, which is such a cool concept. And I've heard it's been wildly successful. And you're also implicitly talking about churn, which I think companies would be wise to think about more or like, you when people leave, why did they leave? We keep having the same problems. for those folks who aren't as familiar with why retention or kind of churn matters for a company? Could you speak a little bit to that? Yeah, and I also have a little bit of a different philosophy, on why or why it doesn't matter. So yeah, 100%. I mean, there's a lot of investment that goes into hiring, onboarding. So the old kind metrics were measuring attrition and we're measuring turnover. So we're measuring employees leaving whether it's by the company's choice or it's voluntary. What I start to like think about now and we're trying to plug into from a hiring standpoint at Humanly is to me, it's a little bit more a company Greenhouse has a term called employee lifetime value. And when I think about it, It's more important to me that you, instead of locking the doors so they don't leave, you're creating such an engaging environment so you get impact in the time that they are here, where someone that stays for four years but is engaged for two years or quiet quitting is the newer term, is not gonna be valuable as someone that just stays for a year and a half but is super engaged, maybe at a onboarding cycle. So I think like if you're able to look at things like engagement, you're, Dr. Brooks Holton at Georgetown has research saying that the worst employee outcome is not a high performer leaving, it's a reluctant stayer. it's hard to measure, but if you can really look at overall employee lifetime value and the impact people are making in the period, are at the company and reducing the time it takes to onboard them and keeping them engaged for the longest period of time. Harder to measure, harder to do, if you can pull it off, it's gonna be more effective than just the retention numbers or just the attrition numbers. But you can only change what you can measure so that it's a lot easier to measure retention and attrition and things like that. So I do think you start there, but. If you can go a little deeper, that could be helpful as well. Yeah, definitely. I think during the pandemic, one of my friends did a, I don't know, filled out a survey at work and found out that like 80 % of people wildly disengaged and like all these performance metrics were not being hit because so many people were burned out or, know, whatever the case may be. But it's like engagement, actively doing the work, you know, moving things forward that the whole company was struggling with. But I think also, right, as you mentioned, like people who were thinking about leaving. Are they doing their best work? Are they their zone of genius, like killing it? Like probably not. And I've certainly joined workplaces where like really bad workplaces where like people were quitting weekly. And just kind of the feeling of that, like, whoa, wait, she left? He left? Wait a minute. Like at least as an employee could be very disorienting or just, you know, as one understands those environments. I've definitely talked to companies that the churn is expensive. Yeah, as you mentioned, these conversations and onboarding people, their recruitment process, taking time to onboard them, they learn the software, they learn the processes, and then suddenly they leave and you have to do all over again, just the number of hours. I heard someone like, maybe you actually know these numbers, but like it can be like hundreds to thousands of dollars to get a candidate onboarded. that, is that true? yeah. Yeah. I've seen a lot of that and I've seen other studies showing that they make up their decision if they're staying long -term, like in the first three to six months and they might stay still, but they could turn into a disengaged worker very earlier. in addition to extending the end date or keeping them longer, there's ways of, increasing impact through getting onboarding right and doing it shorter and doing stay interviews. So you're making sure they're completely engaged. yeah, is very, hiring is very expensive for sure. Yeah, as someone who's not in HR tech You're seeing the gaps in knowledge. Okay, so you are at Microsoft. You're now working more in the data and HR stuff, which they're seeing your passion. You have more interest in this potential. Maybe you're more engaged, frankly, as an employee. What are the next steps after that? Yeah, At that time, I always kind of had a passion entrepreneurship and kind of felt like an intrapreneur at Microsoft. Our VPs were like VCs where you'd go pitch an idea and maybe get it funded or not. So I kind of got some of that itched scratch a bit, but I always wanted to enter the startup ecosystem in Seattle and it continues to have a strong one. So I started mentoring with startups. Microsoft has pretty good. moonlighting policies and ability to do more than just the day job. So I took advantage of all of that. Before I left, in addition to working on HR tech, I wanted to kind of hone my product management skills and work on products. The HR tech stuff was mainly for internal use. So I wanted to work on like one of Microsoft's products impacting millions of users. So I went to Dynamics and I was there for a couple years. And then I decided it was after having my son, who's eight now, my first kid, it was kind of a kick in the butt to think about the impact I wanna make in life. So I was sitting around with him on my lap in paternity leave on the couch, eating a bunch of frozen yogurt and thinking about the impact on the world that I wanna create. I decided then to leave Microsoft, talked to my manager and I, before starting my own company, not just from a skills standpoint, but even just from a financial standpoint, I wanted to go to another more, I wouldn't say like extremely established, but series A, series B company. So ended up at a company called Tiny Pulse. They were in the employee engagement space. So was right in line with the people part of what I wanted to do. data part of what I wanted to do and the product part of what I wanted to do. And they were at Series B at the time and kind of at the stage that at that time I felt would be really interesting for me. I started as a lead PM and then I moved into a director of product role where I managed our design team and product management team. it was a fun, fun ride. I saw a lot, learned how the startup ecosystem was different than big company than Microsoft. So I was at Tiny Pulse for about two and a half years. Okay, nice. And then when did you feel like it was time? Was there more Froyo eating at the next juncture? Yeah, yeah. I should actually, I might need to do that soon. It might motivate of ideas. But frozen custard actually, there's a good spot in Capitol Hill if you're ever there. there's a couple things. One is in my product role was meeting a lot of different people in HR tech I'd form relationships with them outside of work as well. And I really started forming a thesis around something that was completely different than what Tiny Pulse was doing on the employee engagement side, but was still in this HR recruiting context. that was companies that had recruiting teams and were dealing with high applicant volume. They were just so... burnt out by just the volume that there's just an immediate problem to solve. And I found that, you if I think of like Maslow's hierarchy, I was working on like a third tier, fourth tier problem, but people were just crushed and needed their basic needs met. And I felt the basic needs was saving time so that they can have the, in the recruiting side, the front end of the HR process, so to speak, particularly in high volume scenarios. I started seeing that problem through my personal interactions and my network. And then at the same time, when I started at Tiny Pulse, I gave like my two year notice in my interview and said that I told the CEO that I'm going to give you everything I have for two to two and a half years. And eventually what I want to do is start my own thing. He was very supportive and it doesn't always work with everyone, but he was great. when I was starting to leave, actually stepped down from my product leadership role into more of a part -time role. He let me spend some of my time in building up Humanly. So I was able to do that for, a six month period where I was doing both and it was above board and approved by everyone that needed to. So it was a good transition. So I got lucky in that way. then we began the journey of building product. It was started with a bunch of research, having coffee with anyone I could get a hold of that was in talent acquisition or recruiting tech, trying to figure out where, if the pain we thought was a pain was actually legitimate, if people really cared about this problem in terms of high volume screening and scheduling and stuff like that. And then once we felt good enough, I started to raise a little bit of pre -seed money so we could go do it full time and then started selling and whatnot. That's awesome. Yeah. Well, I think it's very fortuitous that you could trust your team and say, this is where I want to be doing or here's the timeframe you got me. not everyone has that luxury, but that's awesome you did. You also mentioned, I think you have co -founders What was that early shape of the team? Yeah, yeah. So I have two co -founders. One is on the technical side, our CTO, and another one is more on the go -to -market and sales side. he was actually at Tiny Pulse as well on the sales team. So that's how we met. I actually met my other co -founder, our CTO, Brian, at the University of Washington. he was a CS major, and we had some mutual friends. when I started Humanly, I was kind brainstorming and rack my brain on who are the best technical minds and folks that I know. he was at the top of the list. I took him to dinner at Applebee's and convinced him to join the company. Okay, cool. Well, and you said from that time to now you have roughly 40 employees. we're about 35. Yeah. because I think some people might be listening to this and like actually building their company and thinking like, who are my first folks? Who are my next hires? Like hiring too fast, hiring too slow, you know, like all that kind of balance. How did you think about those next steps? Yeah, I know it really, it really does depend. And I'll kind of go back to when I say like some of the best product decisions we made or what not to build instead of what to build. Some of the best hiring decisions were made as when not to hire versus just like hitting the gas. having a plan is important, of course, but to answer the question more directly. it was 2019. I was working still at Tiny Pulse. I then left and we went through it. So we did go through an accelerator program. YCombinator in 2020 at the beginning. the team of three. So once we felt enough of an indication as it related to, while I'll go back to saying like we've never quite figured out product market fit, but once we had an indication that people were willing to actually pay money for this thing we were building and we wanted to make more people aware that it existed, that's kind of where we, it wasn't so much hiring a bunch of engineers and going out and building something and taking a risk. was actually, hiring a few people that would help us learn and that was via sales. So going out and doing email campaigns or going out and meeting potential customers. So our early sellers were very much there to learn and give us data and tell us what's working, what isn't, not just to close business, but those were the people we wanted to hire early. And then as we, and we did a lot of things to speed up product development cycles, whether it's, white labeling a third party chat bot to test out and see if people even want to use a chat bot in the recruiting process. So we never felt like we had to build before we sold in the sense of I'm certainly not like telling people it's already built, but it's more like, can we get feedback from you on this concept? Or it was like white labeling a third party so we had something we could sell. and then seeing if people actually wanted it. And then that helped us then decide where we want to invest. So we started with, three folks. We hired a few we can call them salespeople, but it was almost like they were there to research and learn for us. and then we began hiring engineers after that. Yeah. Well, and I also, I'm noticing very much how humble you are because you just told me you raised a series A, you have raised venture capital. You went through YC, which arguably the most prestigious accelerator in the world. do you think of yourself as a low ego founder? usually when I meet people like you, they're pretty full of themselves, I'll be very honest. Well, I really want to win, I want us to be successful and having an ego just gets in the way of that. I like to be, I appreciate the compliment, I've been told that before, but I don't know, I just think of myself as very focused on the problem and what solving for. And I think that that's the way we're going to get to where we need to next. Yeah. Yeah. Well, I think also just it's interesting, you know, you're in a different position speaking with those customers, right? Like, I'd love to hear what your dream customers are. But also, like, noticing the people around you, like from an early stage, you know, that your friend wasn't being treated the same. I think like I have a friend named Josh and my name is Joan. Like we're very few letters off. We both got PhDs from a very, very similar program and his career trajectory versus mine. And like, we share offer letters and stuff. It's fascinating, you know, AB testing -ish. it's, but I think really, I would almost say like being an intersectional feminist in that way to even notice these differences. think, you know, in this modern day, we're having different discussions and vocabulary around. transparency and equity. But I think to be able to see that or even how you're sharing it right now is a different way of thinking about how we build products and who we're building it for. It's just a different lens. Yeah, it was interesting and being a feminist myself as I think everyone should be, but that scenario was really interesting because I didn't go into a lot of detail, but we literally in the same major. We had the same internship, so our resumes probably looked exactly the same. we would be meeting with the exact same panel and she'd come out like literally right after me. we're talking to the same people about the same job. And they very much grilled her on her technical skills and she's a lot more technical than I was. I somehow got a pass there. they would grill me on more like communication skills and just other things, but. It was just so obvious, it was so startling. sets of questions were just not asked to me. So I'm definitely a believer in standardized interviewing, which is a practice here. there's a lot of things you can do to add equity early in the process. One of our measurements, I was telling you, we're starting to measure data and Zoom calls. found even just proximity bias. you are more likely to spend about two minutes more in small talk if you were a Seattle interviewer hiring someone else that's in Seattle versus a different demographic. by the time you got to actual questions, you had a higher sentiment towards them. that part we haven't like totally productized or anything, but it's research we do and we're trying to get better. our entry point right now is at least having a standard consistent pre -screening and screening process across all the thousands of people that apply. That's fascinating. I've been told by my European friends that my pleasantries are way longer than they anticipate. Both my parents are from the Midwest, so I was told to fill the space with a lot of words. Well, as you think about, I mean, this company is growing fast. Series A is no joke. What do you look for in new teammates? And this must actually be very like meta to be hiring new teammates as you are a recruiting company yourself. How do you all think about those next hires? culture add is really important to me. I look at building culture in the same way I kind of look at building products like the... software of old was a disk you install one time and you're good. You might bring in a leader and they install a culture and it's just like that. But now it's very iterative, right? You're adding features to your culture. You have a roadmap for your culture. when we hire people, we want them to kind fit a set of criteria around where we want to go to, not fit a set of criteria around where we're at right now. So I think having a clear vision around what that culture roadmap looks like and the types of components you can add to build it out and get it to where you want to and take feedback, just like in SaaS, you'll take feedback for your customers. In this case, our customers are our employees, our past employees, our future employees. How do we get feedback and iterate on our building that culture. that's important. right now we want people that are here to create impact and they're here for the right reasons, believe in what we're trying to do in the world. those two things as well as the culture add I think are a little more important than just the technical skill. we're in a mode right now where we're very much, I think the seed stage was around ideas and then a pre -seed is more like totally ideas and seed is a little more ideas slash plans. And now A is it's more just execution. Like we have to go out and execute. people that are willing to, have a bias towards action, get stuff done. The main way that like I feel, companies at our stage die is just being distracted or in fighting. we have a very small ship and a set of paddles. So we absolutely have to be rowing in the same direction. Otherwise we're not going to be in a good spot. Absolutely. Well, and as you dream bigger, I'll at least manifest with you. What are some dream customers you'd love to be working with in the future? generally with our customers, want them to be aligned with the future world we're trying to create. So I think things like candidate experience matter a lot to us. Things like diversity, equity, inclusion, and belonging matter a lot to us. When I say we want to make two -way conversations with candidates more efficient and more equitable, on the equitable piece, we want people that... realize like we do that more diverse teams are better teams. so I think mission aligned is one thing, even with our investors, even with other folks we bring on the journey, we want them to be aligned with the world that we want to create. I think they have to see the pain, right? So like they have to have the pain. a lot of some customers might have a need that's more, which is totally fine. It's just a different pain around. they need more candidates. So maybe it's like a sourcing for a more specialized role where they're drowning in them. They're more just needing help in sourcing or using LinkedIn recruiter. And that's not what we're doing. So I think customers that have fit the high volume problem right now, we're doing a lot around healthcare, around financial services, around banking and hitting mid -market pretty hard with our platform. So customers that fit fit those criteria are big and and we've traditionally had a kind of CPG QSR type focus so that continues to be a big part of what we're doing as well. But if I were to just say ideal it would be yeah mid -size finance banking Insurance accounting right now. And I just realized I asked a pretty direct question to someone with a low ego. So that's on me. as you think about folks who may, you know, be really, really inspired by this and be like, hey, that's exactly what, you know, I want to go in this path with HR tech, or I want to build a company that's really solving a problem that I see that keeps coming What kind of advice might you give for people who want to walk a similar path? I think if you're at a larger company, for example, like I was at Microsoft, the hardest thing for me is just starting. So there's a metaphor I give and I don't know like how, if this is actually a thing, but I had read about in India, there's a particular zoo they have there's elephants they have there. there's like some other research on this, but basically they have like this short fence and like the elephant could actually easily step over the fence, but psychologically they see it's a fence and they're not wanting to go over it. So I felt like the stuff that prevented me from taking that first step was very much like these all fences. It's like, well, if I leave Microsoft, what about benefits what if I fail miserably, which some of those can seem like big things, but what you realize eventually it is a really small fence. Like there are ways of getting insurance or ways of like if things don't work out, you'll have some idea of runway so you can come back. And so I think just itemizing your fences and realizing, hey, here's an action plan for each. So taking that first step, I think just doing it. There just seemed to be a mental block sometimes that I had. towards just starting. that is my advice. And then I would say too, there's ways of getting involved without fully committing as well. whether it's mentorship and giving back is a big way of doing that. So I've tried to take on opportunities, mentoring through TechStars in Seattle. It's sad to see the changes to TechStars moving out of Seattle now, That was a big thing for me or mentoring with a group in Seattle called Venture Out. I had no network in the startup ecosystem. Microsoft, great company, but pretty insular. So all my networking energy was spent on meeting people at Microsoft. So doing those things helped a lot. Those are a couple of things I'd mentioned. Yeah, that's awesome. Well, I do think there are a lot whether it's benefits or golden handcuffs that people believe that they do or don't have. I think there's a lot of fear, a of anxiety. I also just think that at least into the future, the way I see it, entrepreneurship is a muscle people are gonna need to flex more and more, especially as the job market changes, all these layoffs. also spoke at University of Washington recently and I was just like, what are the career trajectories for people who are in their 20s right now How are they gonna navigate this jungle gym of an adventure and having side projects or fueling that frustration at work into something really productive could be a path for lot of folks, I would imagine. Yeah, I agree. mean, I think hyper -specialization is kind of a big risk in these days in terms of just learning a very particular set of skills to overuse that quote. I do think ability to learn quickly and adapt and that entrepreneurial mindset allows you to move in ways that you don't if you're just becoming extremely specialized in one particular thing. And there are, this is a very broad statement, there's some professions, like I hope my dentists and doctors are hyper specialized, but in tech specifically, I do think there's a case towards, building skill sets that help you learn quickly, adapt quickly, use data to make decisions quickly and not only focusing on kind of one particular thing. totally hear you. There's some professions that are more adaptable than others. Well, cool. Well, if people want to learn more about Humanly, want to learn more about you, where should they head? Yeah, absolutely. So you can go to our website, humanly .io, find me on LinkedIn, P -R -E-M -K -U M-A -R. you can also follow me on Twitter, just at my first name, last name, tweets. Tweets, awesome. Well, we will have these links in the show notes. Is there anything else you want to share that we didn't cover? Last hand you the mic moment. No, thanks for your time. And like I said, I think a big part in my journey was just starting. So I encourage everyone to do the same. Absolutely. Couldn't agree more. Well, thank you so much for your time. It lovely speaking with you. Cool. Thanks, Joan. Cheers. Oh gosh, was that fun. Did you enjoy that episode as much as I did? Well, now be sure to check out our show notes for this episode that has tons of links and resources and our guest bio, etc. Go check it out. If you're ready to dive in to personalize your AI journey, download the free Your AI Roadmap workbook at yourairoadmap .com / workbook. Well, maybe you work at a company and you're like, hey, we want to grow in data and AI and I'd love to work with you. Please schedule an intro and sync with me at Clarity AI at hireclarity .ai. We'd love to talk to you about it. My team builds custom AI solutions, digital twins, optimizations, data, fun stuff for small and medium sized businesses. Our price points start at five, six, seven, eight figures, depends on your needs, depending on your time scales, et cetera. If you liked the podcast, please support us. Can you please rate, review, subscribe, send it to your friend, DM your boss, follow wherever you get your podcasts. I certainly learned something new and I hope you did too. Next episode drops soon. Can't wait to hear another amazing expert building in AI. Talk to you soon!