
How Ezra AI Is Making Interviews Smarter and Safer (with Ophir Samson)
Loading summary
A
Welcome back to Lead with AI. Today's conversation tackles one of the most consequential gates in modern society, hiring who gets seen, who gets filtered out, and who gets a real shot. I'm joined by Ophir Sampson, who is the founder and CEO of Ezra AI Labs, a company using AI driven interviews to surface deeper signals, reduce noise, and bring more consistency and accountability into the hiring process. This episode isn't just about recruiting technology. It's about fairness, trust, and what it means to design AI that helps humans make better decisions, not faster decisions, as well as helping one of the underdogs get to the top. Let's get into it.
B
Welcome to lead with AI. I'm Dr. Tamara Nall. In each episode, we will take you behind the scenes with visionary leaders shaping the future of AI across public and private sectors. Join us as we explore groundbreaking projects and innovations that are transforming industries and making a real impact on people's lives. Let's dive in.
A
So hello, everyone. How are you? And welcome back to another episode of lead with AI. I'm your host, Dr. Tamara Naul, or some people call me Dr. T. And it is a new year and so many different guests that we have lined up for this year, including the amazing Ofer Sampson, who is the founder and CEO of Ezra AI Labs. Hi, Ophir, how are you?
C
Hey, Dr. T. Nice to meet you.
A
Thank you. Nice to meet you too. And we're so honored to have you. It's because of guests like you and those who are listening that has made us number one in technology on Apple podcasts last year, as well as winning gold in the W3 for guest and interview podcast. So I'm really proud of it, but it would not be possible for you and other guests. So thank you for being here.
C
Thank you so much for having me. Really, really excited for this conversation. I've been a long time listener to your podcast.
A
Thank you, thank you, thank you, thank you. Even though it's a new year, we love to start with the number one question, which is tell us who you are at your core. I feel that every great AI product has to start with some vision, some need. Talk us through that. Take us through the journey of how you came up with Azure AI Labs.
C
Yeah, absolutely. My father was an inventor, and every night he would be kind of awake, coming up with different ideas and building prototypes and manifesting his ideas into real products and putting them out into the world. And that environment surrounded me with, you know, a lot of creativity and science because he was a sister, the core. So the core. I'm a technologist who has had to make a lot of very kind of consequential decisions with very little information, but in the realm of a lot of kind of creative thinking for my career, whether it's been in autonomous vehicles or academic mathematics or hedge funds, or social impact investing, or even magic, because I'm a professional magician in my space, throughout my career, I've hired lots of people, I've built teams, and I've realized that that can be the most consequential decision you have on the impact of your life. The people I've hired, I've ended up spending more time with them than I have with my own wife. So, you know, I think, you know, being really thoughtful about that decision is so important for like how happy you're going to be on a day to day basis. So what really kind of pushed me over the edge with this idea was just seeing how much how AI was just kind of making things so much harder in the problem of like selecting good candidates and who you bring onto the team. And that kind of brought me over to voice because I believe that voice is the most natural interface we have. We speak more than we write. We have been speaking for hundreds of thousands of years. We've been typing for tens of years. And so if you really want to get to understand someone, you talk to them. And so I wanted to kind of apply that to hiring so that we can make better hiring decisions, so that we surround ourselves with kind of better people for, for us and generally lead happier, healthier, more fulfilling lives.
A
Wow, that's amazing. So I'm going to jump because this is like creating my curiosity here. So let's jump to that and we'll come back to some of the other questions. So how does it work? Like if we were to open up the hood, look in the brain, how does it work? How does it use voice to figure out whether or not this candidate is a good one and a genuine at that?
C
Yeah, absolutely. So the way it works is if you're a recruiter, you set up Ezra to, to do an interview for you. And it doesn't just work out of the box. You have to train it in the same way that you would train a real recruiter on your team. You tell it about the job description, which questions you wanted to ask, you talk to it. You literally have a voice. It interviews you and your hiring managers and other people in the team. To understand what is Dr. T really looking for in the ideal candidate for this role. It takes a bunch of information about the company Culture, documents, compensation doc, you can, you can arm it with as much information you want. Then we'll go and talk to candidates and when the candidate has a conversation with, will ask all the questions that you wanted it to ask. But then you'll also ask detailed follow up questions. It's been trained on about 200 leading talent experts, so it knows how to ask detailed follow up questions, how to apply rubrics, how to apply frameworks, how to, you know, when to probe deeper, when to, when to move on to the next question. There's of kind of thinking into what makes a perfect interview and create really great candidate experience. And so under the hood, it's a translation of a lot of cutting edge voice AI into the signals that you really care about as a recruiter to understand is this candidate right for me or not?
A
And it's only using voice, not like, am I looking up, am I scratching my head, am I doing just.
C
No, because people, as I mentioned, I'm a magician in my spare time and so I studied a lot how people act visually, you know, through the art of misdirection and things like that. And people just look in random places, they move their hands and it's very difficult to extract meaningful signal from that. And there's also a lot of regulatory compliance around that as well that we want to be mindful of because it's built in a way that eliminates bias. So we were very thoughtful about what to do. Yeah, yeah.
A
And we'll get to that. Well, that just shatters everything that I think, because I sit across people and I'm like, oh, are they looking up? Are they scratching their neck? And you're telling me that that has nothing at all to do with like genuineness or lying or telling the truth.
C
It's very, it's a very noisy signal. So I mean, there are, you know, misconceptions about when people are looking to the left, they're lying. And I mean, that might be true if you look at it over millions of people on average, but to apply it to like this conversation this time, it's very risky. Yeah.
A
Wow. Okay, now a follow up. So Ezra is conducting the interviews, not your recruiter.
C
That's right. So the one and one thing to say is we take a very different approach to many other companies in this and generally a lot of AI company, you know, let's face it, a lot of AI companies, their core pitch is we're going to replace you, we're going to take jobs away and you can save money and we will take that money instead. That's not what we are about. We're actually not trying to eliminate any human interviews. So the companies that we work with, they're actually doing exactly the same number of interviews as they were before. However, the people that they're interviewing, the candidates they're interviewing, are the ones which are more appropriate for the role. So, for example, let's say you advertise a job, you'll get a thousand people that apply overnight. And now, because of AI, it's so easy for candidates to apply. They write an AI resume to AI generated resume. They send it off, and everyone looks amazing on paper because ChatGPT has written their resume. You've got a thousand AI generated resumes in front of you. You know that somewhere in those, there are 20 people who are really worth your time that you should be engaging with. They're needles in the haystack. How do you find those 20 needles? And so that's what Ezra does. So you'll end up having the same 20. You still end up doing 20 interviews like you did 20 interviews before, but those 20 people are those needles rather than, you know, hey, in the haystack.
A
Got it. So Ezra doesn't just start at the interview where you're listening to the voice. It helps screen people to even get to the interview.
C
Exactly. Yeah. So you would. So it happens before you. Before. If you're the recruiter, you would. You would have Ezra have a conversation that's before. And that means that you're. When you meet the candidate, you know that they have, you know, they. That Ezra thinks they're a good fit.
A
Got it.
C
Yeah.
A
Okay. I was thinking. Okay, got it. So Ezra is screening, and then you, as the HR manager is still interviewing the same 20 that you would without Ezra, but just with, you know, you have those. Those needles.
C
Exactly. So before you, you interviewed basically 20 random people from the thousand. Now you interviewed that. You interviewing the 20 people that Ezra has determined, you know, are a really good fit because Ezra has spoken to all of the thousand people for you. Yeah.
A
Awesome. Okay, thank you for clarifying that. That was good. I was like, sounds amazing. So tell us about, like, that holy smokes moment where. Where a customer, a test user or somebody experienced Ezra and it blew their mind. They were like, whoa, this is amazing. This is more than I thought it ever would be.
C
Yeah. It was a customer of us hiring for a senior engineer for their front end development team, and they had 1200 applicants over the course of two days. Normally what they do is they just look to the LinkedIn profiles and resumes, and really what they have, like, two or three seconds per resume. So they were just looking for keywords, logos. Did this person work at Meta or Google? Did they go to college? The person that that team ended up hiring is someone who, if they were judged by their resume, would never have gotten a look. He didn't go to college. He had no formal degree. He had not worked at a single company that anyone had ever heard of. And so if he was just judged on, his resume would not have been accepted. However, he did a conversation with Ezra just like everyone else, and the recruiter saw that this candidate got a 3.8 out of 4, and so that kind of piqued the interest. Then the recruiter had watched the video of the interview and read the answers, and they thought, wow, this person is really legit. They ended up meeting the person, and that person ended up getting the job. So this is someone who was at the top of the, like, the best candidate out of 1200 candidates. But he would have been completely missed if he was just judged on his resume. So that was a kind of holy smokes. We would have, you know, we would have missed this if. If it wasn't for. For Ezra.
A
What a beautiful story. And I often tell people that I'm like, even when I do my own hiring, I want to find, like, that diamond in the rough. I want to find that person that has this hustle, has this drive, but on paper would never make it. And, you know, under certain people's criteria or whatever.
C
Right.
A
Many. They're like millions of people. Like.
C
Exactly. And Imag had this conversation over text, over email. We 1% of the date. We know each other 1% as well as we do now. Right, Right. So there's just so much richness in voice.
A
Yeah.
C
And. And I. It's so important to give candidates that, like, opportunity to stand out. That one person who ended up getting the role, if he was, you know, again, if you're judged on the resume, he wouldn't have a voice. But he made his case to the company. He explained why he should be selected, and then he got to stand out. And that's just. You could do without voice. AI.
A
Yeah, no, that is. That is absolutely amazing. And. And just changing people's lives and family's lives just by getting the opportunity that he might not have otherwise gotten. That is beautiful. So what about you? You know, as the founder of Ezra, like, obviously you had a vision. It was a wonderful story. Enjoy it. Would love, you know, to. To one day meet your dad, he sounds amazing. But what about something that you experienced with Ezra where it blew your mind? Like, you were like, whoa, did I build this? Like, am I responsible for this? Where you got goosebumps from what it could do?
C
So I'm a technologist at heart, so I'm deep in the engineering of voice AI. I think what impressed me the most was one time where I was testing an interview with Ezra, and I. It asked me a fairly simple question, and I just went off in a completely tangential direction. I start. And I tried to get it to. I actually tried to trick ever into telling me something or saying something illegal. I tried to trick it into asking me about my. Whether I was about to go on paternity leave. And then I was trying to derail it, and I kept interrupting it. I kept, you know, trying to get it to interrupt me. I stopped sentences midway like that to see if it would. Yeah, to see if it would interrupt me. And throughout all of that, it kept the conversation going. It didn't say anything illegal, it didn't interrupt me. Kept the flow very, very polite. And it was just such a cool experience that was. I was like, wow, this really works, right? It's hard. It's really hard. I mean, think of. Think of how much money Apple and Amazon have invested into Siri and Alexa and how frustrating some of these product experiences are. Voice AI is a very, very difficult problem. And so when I, you know, when I had that experience, I was like, wow, we are onto something. This works.
A
Wow, that's amazing. I mean, not only are you in the Ezra team giving people opportunities that might not, you're actually helping employers not get sued from an HR standpoint and making sure that those guardrails are in place and the fact that it can catch those through voice technology is absolutely incredible.
C
So. And another thing we're doing is there's so much fraud right now. So we. About two months ago, we caught our first deep fake candle featuring someone that when. When. When. When a candidate interviews, there's a video of them. You get to watch the video. And one candidate replaced themselves with an AI avatar. So it wasn't actually them. It was. And it was so realistic that we showed it to four different recruiters and they all could not tell that it was an AI avatar.
A
Wow.
C
But we caught it. So, yeah, we're also kind of protecting companies from. From fraudulent candidates who, you know, cause a lot of. Of IP damage to the companies.
A
You know, I always thought, like, how does that work? Okay, so you're going to put Some fake, it's not you, some fake person, some avatar and then you get the job. Like, how do you continue this? Like, what is the ultimate goal? Is this to just see whether or not they can get past it and get an offer and it's like, yes, success. Are they actually trying to get a job?
C
They are, yeah. There's, there's economic, there's nefarious economic stuff like getting the first few paychecks, you know, they get kicked out the laptop, signing bonuses, you know, things like that.
A
The more
C
the other end of the spectrum, you have North Korean sponsored state actors who will find American citizens, who will get the jobs for them and get the laptop and then install malware onto the laptop and then, you know, pass IP or like system access to North Korea. It's not just North Korea, but there's a bunch of North Korean actors.
A
Oh, wow. I always thought maybe it was a group of people trying to do some gaming and see who could actually get through, but no, that sounds serious.
C
It's real and it's coming. I mean, that's the problem with, you know, I'm very passionate, very excited about, about AI specifically. But I, you know, I care a lot about doing this responsibly and ethically and the core value or one of our core values that Ezra is thinking about the impacts of this on society and how do we do this in a responsible way. And the propensity for abuse of. Not just voice but video and kind of avatar and imagery for scamming purposes is just through the roof. So we see it as our responsibility to protect our customers against deepfakes and.
A
Well, and that's amazing. You kind of already went into my next question around like your ethical crossroads and when you talked about working against bias and now the deep fakes and it's a part of your core values. Do you want to elaborate a little bit more? Like, did you start with the ethical compass and roadmap from the beginning and just talk to us about that ethical journey?
C
Yeah, the ethical journey was it started off from what's the problem that we want to solve? We want to make sure that recruiters have an easier job to find good candidates in a world where AI is making their jobs harder. So that, you know, as I mentioned, so that we surround ourselves with good people that we spend most of our time with on the candidate side. It's becoming harder and harder to be a candidate looking for a job right now. So. Yeah, and how do you, how do you give a candidate an opportunity to stand out beyond a Resume. So that was like the starting framework, then the ethical framework beneath that is how do we do it in a responsible way that doesn't try to replace recruiters and doesn't try to turn candidates into a simple number, a simple score. And so yeah, we build a technology very much aligned with those principles and what we have now and the way our customers are using our technology now we have large enterprise customers who are using it and small, you know, who are doing thousands of interviews a month. And small ones, they all trust it to, you know, be, be a good representation of their company. And the ethics of the technology is critical to that. They have to trust that we're doing things in the right way and we're not, you know, we're, we're, we're protecting them against risk. We're improving where we're mindful about the candidate experience. So yeah, do all of that for our customers.
A
Awesome. And so when you think about the future, the future of this world, the future of recruiting, the future of, of job creation, the future of Ezra, what do you see and how does Ezra fit into that?
C
It's a really good question. I, I so, and this kind of goes to my general view on, on AI, there's a lot of, there's a lot of discussion about how it's going to make us all much more efficient and going to save all of our time and everything. But I think kind of realistically it's going to give a lot of. I, I think especially with its ability to create so much misinformation, trust is going to become so much more important than it was in the pre AI era. And we're actually going to go back to more human based interactions. We'll let like the AI do all of the, like the grunt work.
A
Yeah, but then I call them mucky muck.
C
Right, right, right. But we're going to be the ones who will be forming relationship. I think the value of relationships is going to be even more important in an AI era to know who we trust, who we want to spend time with. So I kind of see the future of Ezra as, you know, again, speaking very tactically. You have a thousand people applying. You don't have time or the desire to build a relationship with all thousand people. You want to build a relationship, your human relationship with just a handful, 20 people who you really want to like, understand as you know about, about their fit for your role. And so Ezra kind of does that as you said, that like the mucky work for you to filter them out so that you can Spend more of your time, you know, building human relationships, applying your judgment. You know, Ezra won't make decisions for you. It won't say you should hire this person or not hire this person. That's up to you. But Ezra can give you the space and ability to make those decisions. So. I see. I think that's where the future lies, right? Yeah, yeah.
A
And people, you know, and you mentioned this to people ask, oh, well, AI, replace us, et cetera. And I'm. My thought is that there will always need to be a human in the room. You know, no matter what, there will always need to be a human in the room. So that's important. So our listeners can sometimes be skeptics. They can sometimes say, you know, prove it to me. So if they want to believe in the power of Ezra, or what can they do this week to really experience Ezra?
C
That's a great question. So after this interview, I can share a. I can share a link to Try an interview yourself.
A
Okay.
C
It will be an example interview for a chief return to office coordinator role.
A
I don't know if I want that role.
C
Sorry, sorry. It's a chief return to office enthusiasm. Oh, okay.
A
So I gotta be excited about going back into the office.
C
Yeah, exactly.
A
You.
C
You have to. You. You have to create that enthusiasm. It's a. It's a fake role that we've made, you know?
A
Yes.
C
For example, and you can take a, you know, an interview for it, and it will ask you, you know, two questions, the five minute thing, just for you to experience it so you can see what it's like to be interviewed for. For that role. And afterwards, we can. We can send you the results of. Of of how well you did. Yeah, so that would be a nice way of. Of trying it for yourself and experiencing the technology.
A
Awesome. So y' all will see that down below. Make sure you click on it and. And experience it. So, Ophir, I have a question. Call from one genius to another. And obviously you're a genius here on Lead with AI, and one of our former geniuses posed a question for you, and that is, if your AI had a moral compass, what direction would it point?
C
Great question. It would point towards my fundamental belief that at our core, most people are really good and they have good intentions. Sometimes they have, you know, weird way of. Weird ways of showing it, but ultimately there's something beautiful about everyone, and it's up to us to. It's. It's up to us to find that beauty in other people. And this really kind of goes back to. Sorry to Bring it back to Voice AI. But it, that's one of the reasons why I'm passionate about Voice because it can like find that beautiful thing about everyone. So I think, you know, that that's my moral compass behind.
A
Okay.
C
Yeah.
A
All right, great. So our next section is called Rapid Bonus Fire. So I am going to give you four questions really quickly. Give me the first thing that comes to your mind and if we're intrigued that you can tell us why. We're gonna first start with what is the most overrated tech trend?
C
Less of a trend, more of an idea. That AI is going to make you so much more productive that you'll get loads of your time back.
A
Okay.
C
And that you'll be less busy.
A
Right. We just fill it with other stuff. Right?
C
Yeah. It's like the thing of when, you know, there's a lot of traffic on the road, you make the roads wider, traffic doesn't go down, more people just use the road.
A
Right, exactly. Love that one. First we've heard of that one. What about the most under hyped or underrated AI breakthrough?
C
This won't surprise you. I'm going to say Voice AI. Yeah, it's. It, it's coming. It's coming very quickly, but it's still not. Doesn't. Doesn't have as much attention as, you know, ChatGPT does.
A
Yeah, yeah. What's one book everyone should read?
C
I care a lot. I'm. I say the Social Animal by David Brooks, which is a really interesting story, narrative of ideas of social psychology that once you read it, you're like, oh, I see that over here and over there and over there and over there and suddenly it kind of you. It just opens up so many understandings about how people interact in relationships.
A
Okay, I love that. Now scare us. What's your biggest, boldest AI prediction?
C
It's actually less scary, as I mentioned. I think it will actually bring us a lot closer together. I think there's, you know, there's a camp that believes that it's going to create so much fraud and misinformations and, you know, deep fakes and everything that like we're. Everyone's going to hate each other. I actually think the opposite is going to happen. I think that because of that fraud and misinformation, everything, as I mentioned, we're going to trust interactions like this. Where I know it is Dr. T. You know, there's Ophir. Yeah. We'll kind of go back a little bit to that world where we have more in person interactions, more human to human. Relationships and we offload the like as you said, the grunt work, the muddy stuff to. To AI.
A
All right, awesome. Well, I have definitely enjoyed this conversation as well as I'm sure our listeners. How can we connect with you? What's give us the website, give us all the social media handles both for you and for Ezra.
C
Yeah, absolutely. So I'm best way of following US is on LinkedIn through either following me which is ofeer Samson and the website is Ezra Ezra recruiting dot com. And as I said, I'll post a an interview that you guys can try to get a sense of what it's like to be a candidate being interviewed by Ezra and you can try it out yourself.
A
Awesome. I love that. Well, again, thank you so much for this conversation. I really enjoyed it and I look forward to hearing how Ezra will change the world, particularly for candidates and recruiting managers and that entire industry. So thank you so much for your time.
C
Thank you so much for your time.
A
Absolutely. And everyone, until we meet again, remember to lead with AI.
B
Bye. Thanks for tuning in to lead with AI. I'll see you next time as we continue exploring the cutting edge innovations shaping AI across the public and private sectors. Until then, keep leading with AI.
This episode explores the ways Ezra AI is reimagining hiring and candidate interviews using advanced voice AI. Dr. Tamara Nall sits down with founder Ophir Samson to discuss how Ezra surfaces deeper talent signals, minimizes bias, and identifies “diamonds in the rough”—including candidates overlooked by traditional resume screening. The conversation also covers AI’s role in fraud prevention, ethical design principles in recruitment technology, and why the future of AI-powered hiring still puts humans at the center.
Finding Hidden Talent:
Ophir’s “Goosebumps” Developer Story:
Ezra catches a deepfake candidate:
Nature and Motivation for Fraud:
On Eliminating Visual Bias
On Changing Lives
On Ethical Voice AI
[25:50] Most Overrated Tech Trend
[26:21] Most Underhyped AI Breakthrough
[26:38] Book Recommendation:
[27:18] Big AI Prediction: