AI Super Recruiters: How Tech Is Shaping Sourcing, Placements & the Future of Recruiting — with Steve Lu, CEO of Pin

Benjamin Mena [00:00:00]:
Welcome to the Elite Recruiter Podcast with your host, Benjamin Mena, where we focus on what it takes to win in the recruiting game. We cover it all from sales, marketing, mindset, money, leadership and placements.

Steven Lu [00:00:18]:
I am so excited about this episode of the Elite Recruiter Podcast because I truly believe we are so close to the time where the AI can empower us to become super recruiters. For so long, technology I truly believe has burdened us down. It's almost like going back into like the early 90s where recruiters, all they had to do is like just hop on the phone and work all day and then like stack on tech, stack on tech. And like so much now is just like all the stuff behind the scenes that takes us away from actually doing the recruiting and actually focusing on the relationships, actually making the placements, actually making an impact and actually making money. I think we are so close to that time period where we can focus on all that while behind the scenes artificial intelligence that is made for the recruiting community can help us move forward. So that's why I'm so excited to have Steve from Penn here today. And I was actually super impressed with the product. I actually reached out to them.

Steven Lu [00:01:14]:
I was kind of blown away and I know this world is changing constantly. So Steve, welcome to the podcast.

Benjamin Mena [00:01:18]:
Thank you for having me.

Steven Lu [00:01:19]:
Real quick, quick 30 seconds self introduction before we get started.

Benjamin Mena [00:01:23]:
Yeah, totally. I'm Steve. I am the CEO and co founder of pin.com we are a top of funnel AI sourcing platform. My elevator pitch is basically like think LinkedIn recruiter but with AI on steroids. And so anything you couldn't search on LinkedIn you can. With pin, we like built the ability to search pretty much any kind of experience, skill, tool, anything that you'd like, straight on to the PIN platform and also build extremely advanced filters. Also on top of that, we've built in omnichannel messaging so you can do LinkedIn messaging on our platform, email messaging, etc. And also do automated scheduling across the process.

Benjamin Mena [00:02:00]:
Everything in that whole process that I've told you AI is sprinkled in and helps kind of like you get from your job description all the way to scheduled meetings and you could fully automate it if you like.

Steven Lu [00:02:11]:
And I actually found out about you guys from one of my good friends and I'm actually interviewing him later on this week, so who knows when that would go live. But he's a solo biller that did over 350k in one month and he was like, oh yeah, you know, X, Y, Z, here's the few tools that helped me put together. I was like, oh, let me figure out who these guys are. But anyways, all right, so how did you even end up in this wonderful world of recruiting?

Benjamin Mena [00:02:33]:
So, long story short, pin.com is not my first rodeo. So I built a prior company in 2016 called Interseller. Originally we were kind of this like email outreach tools that you, like, lived on LinkedIn and helped you like email candidates and contacts. But it was originally built for sales. And actually one of our first customers was one of my good friends, Miles Randall, who's actually a recruiter himself. So he worked for Squarespace at the time and he was like, man, dude, I need a product like this. All I do is like send emails, put like reminders on my calendar. I'm like, dude, like, you gotta stop doing that.

Benjamin Mena [00:03:12]:
So what we did was we built another part of our platform, Interseller, into recruiting. And over time, actually in a year or so, we found that like the recruitment side of the business was more up our alley. We were like, the team at the time was much more interested in recruitment. So we built more and more tools towards recruiters. And eventually after five years of doing inter seller, we actually sold that product to Greenhouse to build another kind of integrated product called Greenhouse Sourcing Automation. So that product, Interseller was built more to reach out to candidates, engage with them, do the follow ups, and kind of gave you the tools. So it was all only as good as its operator. And so at the time, during like Covid, I was like, all right, cool.

Benjamin Mena [00:03:53]:
Like, I think it's time for the next chapter. But yeah, also spent two years at Greenhouse, so I learned a lot of recruitment in general on the internal side and external for the majority. And also just kind of realized that AI was becoming a big thing. So after two years there, had a kid, left Greenhouse, and then basically started on a new journey, pin.com, where AI is the forefront of every recruitment technology of sourcing.

Steven Lu [00:04:21]:
I love that I had a kid. Let's go start a new company. Perfect timing, right?

Benjamin Mena [00:04:25]:
It was a little rough in the beginning, just mainly because you have like a six month old and you know, your sleep cycles are really off and startups definitely don't help with the sleep cycles.

Steven Lu [00:04:34]:
But you know, I got my little one's 18 months and we were just chatting before we went live about having a tiny one running around the house. I'm like, you know, sleep when the baby sleeps or work when the baby sleeps.

Benjamin Mena [00:04:45]:
Yeah, yeah, more, more. For me, it's work. I started realizing that you don't know what sleep deprivation is until you start thinking about it in the past.

Steven Lu [00:04:56]:
Yes.

Benjamin Mena [00:04:56]:
So it was rough.

Steven Lu [00:04:58]:
I'm with you on that one. Okay, so you're head deep in AI. You've been working this stuff probably longer than most other companies have been around, working within AI within our space. Like, give it to me real, what does AI and recruiting look like right now?

Benjamin Mena [00:05:13]:
Yeah, so, I mean, there's a lot of, like, companies and products that like, produce AI and you can see it anywhere from literally top of funnel all the way down to like, literally doing interviews for you. So I think, like, we've opened basically a new chapter in the world of computing. Right. There's just so much more products that we can build on top of recruitment, but where I see kind of like the most tedious things and like, this is also why we built like Interstellar back in the day too, was that there's a lot of tedious tasks in sourcing that don't require, like a human interface, like one on one, like you and I are doing on a podcast. But like, our things like, let me schedule follow ups, let me, like, customize the email, let me do all these, like, related tasks that are really important but can be done, like, let's say, on your off time or off hours. Right. And so I think AI can really do those tasks very well versus being able to do tasks like, let me generate, like, I don't know, AI Steve, to then talk to you about, like, let me ask you qualification questions, like, why can't you just do that in, like, the email, for example, to do qualification questions. Right.

Benjamin Mena [00:06:24]:
So I think, like, for the most part, I think there's going to be a lot of noisy products out there in the next, like, year or so, all in recruitment or whatever space that you're currently working in. But I only, I think like, only a few companies get it, if that makes sense. And like, they will be the ones that kind of dominate the space.

Steven Lu [00:06:45]:
I mean, when you talk about companies that get it, I feel like, you know, I see like, of course, like, you know, I have a podcast. I get hit up all the time now by companies and, you know, it's now easier to start an AI company than ever before too. But I really feel like there's a big difference in what, like Silicon Valley is choosing for winners versus what recruiters are. Like, this is what we actually need. Guys like, you just gave them 5 million. We need this over here.

Benjamin Mena [00:07:09]:
Yeah, I think too much. You know, we've seen this with like VC and like investors is that they like kind of like build tech products for tech people if that makes sense. And a lot of the times like they only start realizing maybe in their third or fourth year that like they need to start listening more to their customers. Right. And so I think again like what our team for example has been in recruitment now for like 10 years. Like I hired three or four of my old Interstellar employees and we understand the space so well at this point and especially in sourcing. Right. And so or just kind of like have a 5 year or 10 year head start on any other company that kind of gets it.

Benjamin Mena [00:07:52]:
And so we're building a product with like I guess 10 to 40 plus years. If you aggregate everyone's knowledge of just like being advanced in terms of like we understand recruiters, we know your problems, we want to try to actually solve the things that like are causing you the most headache time are like the most time consuming. And so that's where we are today.

Steven Lu [00:08:14]:
You actually said something I thought was kind of funny because it's actually a conversation I've had multiple times. They were, you know, these tech founders and they were close to recruiting and they're like, well I had this problem when I was actually like looking for a job and so we built a product based on that. I was like, I don't want to pay for that. Like that doesn't solve my problem.

Benjamin Mena [00:08:33]:
Yep.

Steven Lu [00:08:33]:
But it's just like, you know, for any tech founder that's listening to this, like talk to the recruiters, figure out what we actually need, figure out how we can move the needle, figure out how we can do these things.

Benjamin Mena [00:08:46]:
So yeah, exactly.

Steven Lu [00:08:48]:
Out of curiosity and we get, we're going to do a bunch of stuff on AI like you said, the next six months we're going to have lots of noise coming at us. Why?

Benjamin Mena [00:08:58]:
Yeah, because ultimately the Companies that build AI models like OpenAI, Google, Gemini, Anthropic, basically are now releasing models almost on a bi weekly basis or whatever updates that they do to those models. And so they have added functionality like improved like rule following. All these different things that help kind of like us. Like for example, I'm an engineer, so helps me kind of like deliver a product faster. Right. And so there's just going to be much more new capabilities to unlock essentially that like was either impossible yesterday and will be like possible tomorrow. Like here's a very good example. Four years ago it used to take us like you know, a month's worth of time to read text on a piece of paper from an image, like let's say handwritten text.

Benjamin Mena [00:09:48]:
And so we call that like ocr. And it just took a long time as engineers to build. With today's AI models I could integrate OCR within like a fraction of an hour. That's just kind of like how an advanced essentially AI models have become. And so I think like over time we're going to start seeing more like better voice recognition. I wish like this thing here, like I'm not going to say it like because it starts with an S and you spell it with iri basically is terrible at listening to what my words are or interpreting it. But I do think in the next year or so like that thing is just going to get like thousand times better at listening. Right.

Benjamin Mena [00:10:26]:
So yeah, just like over the next few months there's just going to be more unlocks of things that people and engineers can build. That was much harder to do back in the past where you had to hire like multimillion dollar engineering teams. Now you could just do with AI.

Steven Lu [00:10:40]:
Models like what are some of the things that you think will be like coded faster to be able to like solve recruiters problems?

Benjamin Mena [00:10:48]:
Yeah, I think like resume parsing is one. So like it used to be that like, let's say whatever ATS you're using would have to like take in the resume, destructure it into like a format that a computer can read. AI models now can do that almost in a matter of like seconds and probably a lot cheaper than you like think also that's a big one. I think over time it can start asking like deeper questions and better at writing. For example, like let's say I have a job description, has a few requirements and I read a resume now and the job description to be like, does this person match those requirements? If it doesn't, I can actually create questions to the candidate saying like, hey, I noticed you mentioned typescripts. Do you also have react experience? Right. Like it can kind of naturally flow in the language that like we write. And so candidates are more enticed to respond to you once AI models get better and better at recognizing and matching your like, your tone.

Steven Lu [00:11:53]:
So with all this AI that's happening in the recruiting space, like I don't know how to properly say this, but like how does a recruiter stand out with the AI slop that is out there? I mean on top of the slop, like also the really crappy 1990s emails that still get sent every single day.

Benjamin Mena [00:12:10]:
Yeah, I think the best thing you can do As a recruiter today is like really sharpening your skills as like a relationship builder. I think like the best recruiters I know are like always talking to like clients and always talking to candidates and being kind of like, let's say the pillar of like the community, for lack of a better kind of word. So like an example here is that like I have a Rolodex, I know everyone in this like for example industry. I know exactly what they're looking for. I know exactly how the industry is moving. So ultimately like, you know, like let's say software engineering recruiters, they're more like the best recruiters, know like what AI does. They know what real AI companies are and not are. Right.

Benjamin Mena [00:12:55]:
And so they're more relatable to the industry and know how to like basically talk to an engineer and also talk to the company. Right. And so like when a company comes and says like, I'm not really sure what I need to hire, but like, this is kind of what I'm looking for, that recruiter basically will be able to say like, yep, I know exactly what you're looking for because like I've done it a thousand times. Right? You're supposed to be like the expert. And so being more of that like industry expert is going to be more valuable than like help me find a hundred resumes in the pool, if that makes sense.

Steven Lu [00:13:29]:
That definitely makes sense. And I'm already seeing that just the way like the leveraging of the social media and all these other things that what used to be like big go to's for the aero tax of the world you can now do yourself.

Benjamin Mena [00:13:42]:
Yeah, yep, exactly. I think like, if I would explain like, like really well what AI can do is imagine if you had a team of a hundred kind of like doing that role and they didn't interface with either a client or a candidate. Like you can now do that with maybe like 10 people or less. I think the majority of, let's say, for example, like, you know, there's a lot of products companies that recruiters buy to like kind of do their sourcing. I think that's going to be a lot less fairly soon, even though that part is much cheaper. You're going to just get way more consistency out of AI versus like your virtual remote assistant consistent in the Philippines or Vietnam or wherever in Asia. Right. And so you're just going to get a lot more consistency and a lot more rule following in terms of like, here's what you need to source, give me more of these candidates, less of those and AI is going to be able to be like, great, here's everything I can do, right? And so I do think, like over time the need for human labor offshore is going to be a lot less.

Steven Lu [00:14:45]:
I've been saying that for a while, especially if the cases of like, I know so many recruiters that just like they walk in, they're VA or the team member has like built a list for them to start attacking. I was like, I think I was just like a week ago, like, yeah, give it like four to six months. And either your VA needs to be a supercharged recruiting engineer or the AI is going to take it over.

Benjamin Mena [00:15:05]:
Yeah, yeah. Like for sourcers specifically, I think like, you know, the best way to future proof yourself is it's not a numbers game anymore, it's a quality game. And so I think a lot of the, you know, a lot of recruiters that we speak to almost on a daily basis still think, think it's a numbers game. If I put like 10,000 people into my funnel and I have a response rate of like 1% or less, I will get the numbers that I want. But I think over time AI is going to make it harder to do the numbers game. And so you need to start playing the quality game. And so it's just a matter of like a function down the funnel, which is like, if you, if you lower the amount of people but increase your quality, your response rates are still going to like going to skyrocket, which means you're still going to get like the numbers that you want. So yeah, it's just more like, can I help you perform that quality without you spending the time? That's really the key.

Benjamin Mena [00:16:01]:
Kind of like what we're doing here at pin.

Steven Lu [00:16:03]:
All right, so before you like jump into like what PIN can actually do for many recruiters, the elephant in the room, doom and gloom. Like, is AI going to get rid of the recruiter?

Benjamin Mena [00:16:14]:
I actually think it's going to be the opposite. And here's why. I think we're going back to full circle to like boiler room style interviews where everyone's like in a waiting room and then being interviewed by the recruiter every five minutes. Right. And so why do I think that? I think we're in an arms race. And the arms race is the candidate has AI and recruiters have AI and the arms race just starts becoming like, I'm just up leveling one by one. Candidates do this. Recruiters have to combat it.

Benjamin Mena [00:16:44]:
Like I start like putting like an AI virtual like avatar for myself to interview with the recruiter the recruiter starts doing it. Right. We're just in an arms race right now. And I think it's primarily driven by one, AI models getting really good and two, actually like remote hiring because we're so fixated on like doing zoom hiring because it's so much more efficient, if that makes sense. So, like, I think we're just going back to the room where like, I can't trust, like, if you're a real person, unless you know you've got some like, other background checks that I've done or reference checks, et cetera, like, we're getting back to that, like boiler room style, especially for like incoming applicants. I think it's just going to be more like you have to be in person in order to interview. Whether it's in the beginning of that stage or the end of your interview stage. I think recruiters are going to be way more important in the next year or so because I think AI is just an arms race today.

Steven Lu [00:17:41]:
2 thoughts on that. Real quick. First of all, I just saw on LinkedIn, a CEO of a big tech startup was talking about somebody needs to invent interview hubs that are in person.

Benjamin Mena [00:17:52]:
Oh yeah, I think I saw that too. Yeah.

Steven Lu [00:17:55]:
Because of just how much you were telling me offline that it's hard to tell now what's a deep fake. But how do you tell what a deepfake interview is? How do you break out the AI?

Benjamin Mena [00:18:04]:
Oh, yeah. So one of the, I think one of the funniest things is that like, for live, like video models, they're not good at kind of detecting kind of like the face. So they look for like eyes, mouth, ears, et cetera and like kind of a generic face. So if you ask them to like wave their hand in front of their face, you'll start seeing like their face changing because it can't detect, it doesn't have like memory, so it doesn't know a hand is in front of a face. So you'll start seeing essentially like their face change very intermittently. And then also another tall tale sign is they, they will ask to not do that. They like don't want to do it, even though it's like the easiest thing to do. And you'll see it like night and day.

Benjamin Mena [00:18:45]:
Also another thing to do is have them turn off their virtual background because that's like another item that kind of like AI does very, very like, well. So keep in mind that, like, who knows what's going to happen in a year. I do think, like, the current combat is just hand in Front of the face, really fast, like a hi. Just like saying hi and things like that. It can't do very well. But I have seen models that are very, very good. So like, again, I think like today those models like require like a lot of processing time, so it can't be done live. But I do think like in a year's time, those models will make.

Benjamin Mena [00:19:22]:
Make people look like crazy. Good.

Steven Lu [00:19:24]:
Okay, so I got a question. So anthropic CEO has been talking about this OpenAI. You know, Sam's been talking about this. At OpenAI, I play with operator, which I think has so many capabilities, but it really runs around like a drunk toddler. It's like giving my kid my 6, 18 month old, like the mouse and it gets lost constantly. But there's all this talk of like these AI employees or these true AI agents. And I know OpenAI has been talking about charging like 10 to 20,000 for these. Are we going to see in the future these like true AI recruiters that are almost like AI employees?

Benjamin Mena [00:19:58]:
I think overall you have to think AI model. It's like there's a realm where it's like you mentioned toddler, right? So there's the toddler's level of information where it's like, okay, I get it, trying 20 different things to get to the right answer. But there's also kind of the opposite problem. It's kind of like a bell curve. So I think experts live here in the middle of that bell curve, which is like, oh yeah, dude, you just need to do this. Just press this one button, then do this, then that. Right? Because we're TR on it. And then there's like, basically there's too much information where the AI model knows too much about everything.

Benjamin Mena [00:20:34]:
So then it goes like, wait, I could try this or maybe I should try this, or maybe should I do this? Because like, it's basically reasoning out all the different things it could possibly do. Because again, it's like it knows too much. And so like, there's a realm here. I'll say, like there's this bell curve where you have to be somewhere over here to basically be very efficient at like what you need to do. But let's get back to like kind of like the AI agent question. I think, like, I've, like, I have a few friends, engineering friends, especially one. Actually he is basically creating an engineering team based on AI models. Very specific.

Benjamin Mena [00:21:10]:
So like he has a QA engineer, he has like two software engineers, an engineering manager and a product manager. And basically they're in a Chat room, talking to each other, performing their roles. And funny enough, it like operates like pretty well. I think for the most part. Like, it's crazy to see the conversation that it has, especially in that context. And it does actually like, fairly well in the sense of like you describe it kind of like a product. You're the executive describing, like the product that you need to build and helping it guide along. And it's just doing the engineering process.

Benjamin Mena [00:21:43]:
It definitely is like getting closer and closer. But I still think we're like years away before, like we have like an AI recruiter. But I think the key thing is that there's still specializations. Just how I mentioned, a QA and software engineer. There's going to be a sourcing agent, for sure is going to happen. There's going to be a scheduling agent, for sure is going to happen. Right. I think currently, right now we're going to have these micro agents everywhere that replace certain responsibilities of a recruiter's role.

Benjamin Mena [00:22:19]:
But a full kind of AI recruiter, I still think it's years away, but I'm very optimistic that we probably will get there soon. I think today the way we should think about it is we should start ourselves basically building AI agents that operate those tasks for us, rather than thinking completely just do my job, if that makes sense.

Steven Lu [00:22:43]:
That makes sense. You know, you mentioned about your friend that has like these like engineering AI agents, like, can a recruiter like Vibe code their way into like a product now?

Benjamin Mena [00:22:54]:
Yeah, I mean, I think I have my friend texting me right now being like, can I do this? And I'm like, yeah, absolutely. Just go to Anthropic. I think they just like launched Claude codes on their like, max product. It will do very well. Like, right? So you say like, hey, I have this CSV that's really unclean, has like bunch of LinkedIn URLs and it's just going to like write code for you to put process or like, let's say you're like, oh, I've got like, I collect the CSV from my va, like a thousand candidates and my product has an API. Can you just import them automatically? Right. You can Vibe code your way through all of that. If you want to build like, like a basic website, that's even easier now.

Benjamin Mena [00:23:34]:
So you know how we used to have like Squarespace and those products and tools? I think you don't need those anymore. I think you could just ask like Claude to just build your website. Right? Like, you know, let's say you're starting a recruitment Agency, you need like a homepage. I think you can do it. Like, I think a recruiter can vibe code that and like, give it a day or two. I think for the most part, a super awesome website. Yeah, exactly. Like anything that you want, you just have to describe it and you have to give it feedback.

Benjamin Mena [00:24:02]:
But I think you'll get pretty close to exactly what you want.

Steven Lu [00:24:05]:
That's insane. As I go viper code myself a website tonight.

Benjamin Mena [00:24:10]:
Dude, you should try it. I mean, like, I've been vibe coding a lot of like, smaller things to like, automation things in my home. And it's just, it's good. Like, like turn on the lights at a certain time when the luminosity is this or the weather is that like it. It does phenomenal job.

Steven Lu [00:24:25]:
All right, so questions about PIN. So, like, for an average recruiter, what is PIN?

Benjamin Mena [00:24:30]:
Yeah, so we built basically like an AI kind of like product with AI first, right? So we came into the light of AI, but we basically built LinkedIn recruiter, basically the sourcing aspect of it merged tons and tons of data with it, like, you know, job, like, past job descriptions, funding round information, all of this other relatable information and merged it into a profile. And then we had AI basically just read every single person's resume. That gives us the ability to know, like, what experiences do people have? What experiences are people looking for? Right. And so we're able to kind of match with extreme precision it's exactly what you want without you describing, like, or finding keywords. For example, the old school way is like Boolean, search keywords and all that kind of stuff. The new way is just describing exactly what kind of person that you're looking for. Like, I want these skills, these experiences, like migrated databases, built apps from scratch from 0 to 1. Like all of those things you can ask.

Benjamin Mena [00:25:31]:
And our AI is able to prove that because, like, it's read every single person and can deliver essentially like those people to you. So, long story short, we built like LinkedIn recruiter with AI on steroids. Also built kind of like everything that like interseller had as a product, all the tools that recruiters need. So we have email sequencing, we have email finding. You can do omnichannel messaging, like on LinkedIn connection requests, et cetera. You can do all of that on our platform while also having all the AI that you need along the process. So we can write you that email sequence. We can customize each and every email message that goes out to a candidate based on their profile resume.

Benjamin Mena [00:26:11]:
The companies they've worked for we can even do all the questions so like think like they come in, they ask you a question, we can answer back in the same exact tone as you. We can do all these things to just basically get to that end result which is like you having that first interview with that person. Right. So we do scheduling also. We do it all via email and text versus through like a cal.com to make it extremely personalized. And so the end goal is like if you get it, everything personalized without it making it feel like I'm just like waiting in line for the process, does that make sense? Like the candidate feels phenomenal, right? They're more inclined to respond back to you, they're more inclined to interview. And that's the goal of PIN is just like try to make everything as like human in like quality as possible. So that way that candidate is speaking to you and you obviously can close them towards that end process.

Benjamin Mena [00:27:08]:
So we've built everything top of funnel sourcing that you could possibly ever need with AI, all the AI tools and agents that you need to and very much like give us a job description, we'll schedule you interviews.

Steven Lu [00:27:22]:
So it's great on the sourcing side when, if you guys want to, if you. Good laugh. I kind of shared one of the candidate feedbacks on LinkedIn and somehow got way too many comments. I think I spent seven hours responding to people. Can PIN help me with the business development side at all?

Benjamin Mena [00:27:37]:
Yeah, same kind of like way you ask like PIN about like hey, I'm this specific industry recruiter. You can actually do the same level of search. Say like I'm looking for CROs that I want to pitch to. And like in this specific industry you don't need to mention the companies anymore. We know every single company that's out there and you get like again a giant list of candidates. And then at the same time basically you can switch our email lookup system from personal work emails and you can also get like mobile phone numbers too. So all of that, you get all of that data, you can run a different type of sequence which is like a sales like sequence all on our platform. We can handle everything and you can even change the messaging up.

Benjamin Mena [00:28:20]:
So like let's say like you're not pitching a candidate anymore, you're pitching a client like a company. There's this box where you can like change how the message is written and then basically get more professional or more, more salesy however way you kind of sell. But we can get very, very close to like how you do your client development. Today.

Steven Lu [00:28:37]:
That's cool. Like, but you. You said something like, you guys know every company out there. And I'm like, man, can you guys, like, help me market map?

Benjamin Mena [00:28:43]:
Oh, yeah. I mean, like, we. We have some, like, behind the scenes, like, developer data that tells us, like, where your employees are joining from and where your employees are going to. Right. We have, like, tons and tons of data like that. We can pretty much build, like, crazy levels of maps. Before, it used to take a while. Now we just like, AI, here's our data, Go run this report for us.

Benjamin Mena [00:29:09]:
And it's pretty. It's crazy. Like, I think the technologies I'm most impressed by is like, AI models are great, but I think maybe like this close second is like, big data warehouses and databases. Like Snowflake has gotten really, really good over the last year. There's another product called Clickhouse, which is phenomenal. Those, I think, are like, my close second, right behind AI, the level of data that we're working with these days. And like, I hate to say it, but I think, like, Snowflake and Clickhouse are like the shovels and the gold rush, if that makes sense. They're selling kick axes and shovels where, like, AI models are obviously doing really, really well, but those guys are selling the pickaxes and shovels.

Steven Lu [00:29:48]:
To me, I mean, everything runs on data.

Benjamin Mena [00:29:52]:
Exactly. Yeah. Yep.

Steven Lu [00:29:54]:
I mean, I was sitting there, you know, this is pre, like, OpenAI and everything that happened with them. But I was, like, studying what was happening with AI in China, and they were talking about all the different, like, social apps they were using for their data models that were there. And they were like, America can never keep up. And then, of course, you know, OpenAI AI changed that.

Benjamin Mena [00:30:11]:
You know, we have Deep Seek and those models now. Even back in the day, like, even before this whole, like, AI rush, there's. There's an institute called, like, bai. It's like Beijing Institute of, like, Artificial Intelligence. Like, the models they create are, like, pretty insane quality before even, like, Deep Sea kind of came out. So I don't know, I think we should be, like, slightly worried about, like, China also just, like, building models that are, like, phenomenally good or, like, very, very good or close to being like, let's say, like, us models are here, like, at 100%. I think they are getting to, like, 90% without, like, one. They don't even have the access to the crazy hardware that we do in the United States.

Benjamin Mena [00:30:53]:
They have, like, liquidated versions of, like, Nvidia hardware. So they're able to deal with Like, a lot. Like, they're able to do a lot with a lot less of that. So I'm slightly worried, to be honest. That's my personal opinion. And they just don't have, like, they have other crazy politics, you know, like, you can't mention the president or anything like that.

Steven Lu [00:31:12]:
Details.

Benjamin Mena [00:31:12]:
But we have. We have our own version of it, which is like, most of the AI models in the beginning had basically, like, people evaluation, like, recruiter, like, stuff trained out of the models. So we had to train it back into the models when we first started. And now, like, over time, people realize that, like, wait, this isn't as bad as we thought it would be. So, yeah, back in the day, man, we had to train our own models or fine tune our own models with human data because it just refused to, like, evaluate resumes and things like that. Like, not well, at least.

Steven Lu [00:31:44]:
I mean, I haven't checked it recently, but I remember, like, trying to do stuff and they're like, nope, I cannot give you an answer. Literally would tell me that. I'm like, but you could do so many other things. What the hell?

Benjamin Mena [00:31:52]:
Yep. Yeah, it's getting. It's getting better. Like, the guardrails are definitely being loosened up a little bit. So it's definitely gotten better in the last, like, month or so. Two months or so.

Steven Lu [00:32:02]:
What's the one thing that you're excited about with the future of AI?

Benjamin Mena [00:32:05]:
With the future of AI? I think it's like an efficiency thing because, like, again, I'm an engineer. I love, like, building efficiencies. I do think in the engineering space I can build, like, a product that typically takes like, a team of like, 15 people. And like, just to put in context, my engineering team and the product we built today is three and four, so four, including myself. We're really small and lean team, but we're able to build, like, three companies worth of products. Like, it's. I think that's kind of like the future where I'm like, wow, we can really get kind of, like, some efficiency gains. I do think, like, it upscales the United States.

Benjamin Mena [00:32:47]:
I think, like, it also, like, solves the offshoring, like, knowledge problem. I think the more we use AI models as a population, the better we can kind of, like, upscale our skills. Right? We can. We can do a lot more. We can go broader in terms of, like, what we're able to do in our capabilities. We don't need to just be like, the best interviewer or the best sourcer. You can do all three at the same time. I think it just upskills everyone.

Benjamin Mena [00:33:15]:
But there is kind of some downsides which I think like, is kind of like harmful for future kids, if that makes sense. So I'm weary of those things and like, have a vigilant eye on that. But I do think I have a lot of optimism in AI models, really solving like real world problems.

Steven Lu [00:33:32]:
That's awesome. Oh, and one quick question. I've had this asked me like six or seven times. How are you guys with EU data?

Benjamin Mena [00:33:39]:
Yeah, so we all have European data. Most of the block, I think, like only one country we don't have, like. Yeah. And the best part, yes, our AI models can read any language that you want. So if you say, like, let's say I'm looking for this specific skill or experience, it can read French, it can read Spanish, it can read practically any language that you want, German, you name it and like, it will translate the profile for you. And in pin, you'll see kind of like exactly who you're looking for. And so like, I actually think we have more profiles than people think because typically on like a LinkedIn recruiter search or any other search product, you're not getting all the European profiles because you're not able to search French words, for example. Right.

Benjamin Mena [00:34:28]:
But doesn't matter. PIN can read any language and interpret it like exactly the way you need it. So I actually think we have more coverage than LinkedIn does because again, like, we have more non English profiles that most people cannot search.

Steven Lu [00:34:42]:
So this is like non recruiting related. It's something I'm excited about with AI because, like, we're planning on spending some time in France this year. And like, I see that you're wearing AirPods and it looks like we're supposedly dropping this fall, like auto translation through your AirPods for like ton of languages.

Benjamin Mena [00:35:00]:
Yeah, you know, like I'm, I really want that for sure. My wife's Chinese and all of my friends speak Chinese except me. But I can, I can listen to Chinese. I'm really bad at speaking it. Here's the thing though, I'll see it to believe it because I think Google has touted this for like years and it's like no one uses it, the quality's not great. And so like, I'll see it to believe it, if that makes sense. But I really want to see it because I think that's just going to improve just like communication stuff, skills, like across the board.

Steven Lu [00:35:32]:
I have a, a pod that I take with me for certain places and my wife was having a conversation with somebody in French and it was giving me like full on translation on the back end. I was like, this is amazing.

Benjamin Mena [00:35:43]:
Wow. Yeah. Crazy. Yeah. I mean, I believe it just because, like I see it in our AI model usage today. It's just like, it's insane at like getting things accurately.

Steven Lu [00:35:54]:
Well, before we jump over to the Quick Fire questions, then they don't need to be quick answers. Is there anything else that you want to share about AI, the future of recruiting? PIN?

Benjamin Mena [00:36:04]:
Yeah. So I think at PIN, we're designing this brand new search called V3 and man, the results are insane. I think we are actually building the next generation of search with the current AI models. You can practically do searches that you could have never have done before in the past. An example is like you could do Nose has been working with REACT for the last seven years, for a minimum of five years. And you can do those really advanced, like kind of crazy technical questions. Like, I'm looking first. Like, I think traditionally like a search that's been hard is people with like TS or SEI clearance.

Steven Lu [00:36:47]:
Like, like, that's my space. Yep.

Benjamin Mena [00:36:51]:
Dude, we can find like hundreds of thousands of people now because we can interpret like, if this is like, if this company that they work for is currently like a government contractor, like it knows all of this stuff. So you don't need to explicitly say that I am like TS cleared because like a lot of candidates are really afraid to like put that on their LinkedIn. But we can interpret it like, if you're historically part of like a government contractor, we can be like, we assume this person does. And the one thing I'll just have to mention to every recruiter out there is like proof on a resume. Doesn't mean that they don't. If they don't list, it doesn't mean they don't have it. So you should definitely reach out to ask. Right.

Benjamin Mena [00:37:30]:
It's a conversation starter. And I think a lot of recruiters are still kind of afraid of messaging candidates because they don't want to look bad on behalf of their client. So again, like the job of a recruiter is to engage with people and get the information you need. Like, if you're not doing any of that, basically like AI is going to replace you in that world for sure.

Steven Lu [00:37:53]:
All right, we're going to have another clearance conversation once we're done recording.

Benjamin Mena [00:37:57]:
Cool. Yeah, sounds great.

Steven Lu [00:37:58]:
Yeah, happy to like 15 years in that space.

Benjamin Mena [00:38:02]:
Awesome.

Steven Lu [00:38:04]:
Well, anyways, jumping over to the Quick Fire questions and like I said, they don't need to be quick answers, but love to hear your perspective on some of these things, you're at an event, we'll just say example, like recfest and a recruiter walks up to you. This is their first year ever as a recruiter and they're like, hey, you're doing a lot of AI stuff. You've been working with recruiters, now your third company, working with recruiters. Like, you've seen a lot. If I want to be successful in this career, what piece of advice would you give me?

Benjamin Mena [00:38:31]:
Yeah, in your first year, I would say probably two things get really, really, really, really good with AI. And number two is build your tech stack that you think is going to like basically multiply you by 10. I think the biggest thing that you can do early on, especially in this day and age, is like instead of trying to do things like manually one by one, they're still kind of like, start figuring out very early in your career, how can I get AI to do these tasks? Because you're at this point in your career where you haven't seen everything but you're starting to like learn a lot of the like, oh, my boss gave me this task. It's really tedious. It is like the most tedious thing. I just repeat the same thing over and over again. Start figuring out how AI can do that. Because if you know how to do that, you're better than your boss almost at that point.

Benjamin Mena [00:39:27]:
Being able to figure those things out and being exposed to that is super, super important. So like one thing also on the AI side is like the second piece of advice that I would say is kind of like slightly the opposite, which is like understand why you do what you do today. Meaning that don't try to just like vibe code your way through recruiting, if that makes sense, and not understand like why things are done certain ways. I think the most important thing you can do is understand how things became to be like the process. Like all the different things that like we've developed to make things more efficient. Understand it and then don't be afraid to change it. Right. Because AI can help you in those parts.

Benjamin Mena [00:40:08]:
But if you're like stuck in the process and you don't understand it, you're going to get lost when you start putting AI into the process.

Steven Lu [00:40:16]:
Same question, but for, we'll say an old dog like me that's been around the block for like two decades now, what advice would you give to one of those recruiters?

Benjamin Mena [00:40:23]:
Ooh, like the decade long recruiters, I think in the talent space, just like, you know there's going to be the top 5% of recruiters that bill like you know, 400,000 per year. Right. I think if you think AI can't do what you do, you're going to lose in the next five years. I think the decade long recruiter needs to start figuring out what can AI do to improve me. To like, let's say I'm billing 500k per month. That's crazy by the way. And if you like go up to like a million dollars, like you're just bridging the gap so much further. Again, I don't think AI is a replacement for you.

Benjamin Mena [00:41:07]:
I think AI is an enhancer. And so what that means is there's going to be a pool of like recruiters that are using AI and guess what's going to happen? They're going to upscale, which means they're stealing business from you. So I think if you are saying that like AI is like never going to replace me, I'm never going to use AI. I think you need to get onto the train because like that train has left and like, if you wait another like six months, that train's gonna be long gone.

Steven Lu [00:41:36]:
All right, So I actually, I think like AI is gonna, like, this is just me personally, like, I think you're right about that AI, but do you think we'll see a time where a solo biller or like maybe like a super small team can start cranking out 500k months with AI?

Benjamin Mena [00:41:51]:
500K a month. Let's do the math on that real quick, right? 500,000amonth. I mean just say like, let's say your 200k annual salary times like 20%, right? That's 40. So you need like 12, right? I've seen people do 8 to 10 without AI, so I fully believe you can. With AI, I think you can actually do even more than that. So I've had friends, the recruiter friends who I like know are killers. I've done like half million dollar months and I'm just like, every time I hear that, I'm always like, man, I'm in the wrong. Like if I did, if I did that every single, every single month.

Benjamin Mena [00:42:33]:
Obviously it's not every single month. Like you can sandbag it a little by like, you know, pushing people off into the next month. But yeah, I fully think you can hit half million dollar months with like a solo recruiter by yourself for sure.

Steven Lu [00:42:46]:
All right, cool. What is one book that has had a huge impact on your career?

Benjamin Mena [00:42:52]:
Ooh. So I'm gonna let you into a little secret. I don't have never read a book in the last 10 years. I don't like. The one key thing that I think I fully kind of believe in is like, again, I'm just gonna preface this. I'm an engineer, I'm a tinkerer. So I like when someone tells me, like, hey, this is the sales process. And I'm like, okay, why is it done this way? Like, and then I experiment with it, and then I learned from basically examples and things like that.

Benjamin Mena [00:43:23]:
So I tend to try to avoid just, like, reading the book on things and, like, more like try the book if that makes sense of like, what typically people have done. And I think the best, like, alternative, which is also kind of like the most expensive alternative to a book, is I hire a lot of advisors around me on specific things that, like, I don't know about. So that way I can ask them tons of questions and learn from them. And that's like the alternative, like, the most expensive alternative to a book, which is, like, I have the experts around me that tell me the things that I need to do, but the things I read is resumes, code and code documentation. So the problem is I spent like 10, 12 hours on a computer. So, like, a book is just really hard to, like, just jump into because I've already been staring at text, like, for the last 10 hours. But, yeah, that's a little secret of mine. I have not read a book in a very long time.

Steven Lu [00:44:20]:
For you, like, you know, you're heads deep in tech, what is your favorite tech tool that helps you out?

Benjamin Mena [00:44:26]:
Cursor, without a doubt. So cursor is a vs. It's basically a VS code extension. It's like all, like, VS code is by Microsoft. It's the editor of choice for, like, many, many, many engineers. And cursor is an extension on top of it, which is ultimately a kind of like, AI kind of like, version of coding for advanced engineers. What it can do is basically like, I can probably write four times as much code with cursor than I have ever before in the past. Like, I think the one key thing is you press tab and it autocompletes the code.

Benjamin Mena [00:45:02]:
It does a very, very good job at doing it. You just need to be, like, conceptually in the right path and can conceptually get it correct. What typically would take like, an hour. I could probably, like, I just did it this morning for. For somebody. I was just like, five minutes. Five minutes code for something that he thought was gonna take four hours. And I was just like, yep, did it in five Minutes.

Benjamin Mena [00:45:23]:
I'm not a big advocator for anything with, like, developer tools or like that, but Cursor in the last two years has definitely changed my mind on things. And usually I'm that guy when someone goes like, I'm like the old guy in engineering. I'm like, oh, another one of these. Like, I go like that whenever I hear, like, new tech products. But Cursor definitely was like, wow, this is like a night and day difference of, like, how I can write code. And funny enough, like, a lot of people take my suggestion about that. So I've gotten like, full, like, thousand person engineering teams using Cursor. I wish they gave me a referral fee because, man, I put it in, like, really, really large teams.

Benjamin Mena [00:45:59]:
I was like, man, I wish I could get at least 10 of that.

Steven Lu [00:46:03]:
That's crazy. What is one of the biggest failures that you had to, like, work through?

Benjamin Mena [00:46:08]:
Hmm, Biggest failures? Oh, man, there's probably too many failures. I think one of the, like, the biggest challenges or like, like, maybe this isn't a failure, but a mindset failure, if that makes sense. It's like, I think sometimes for myself, there's like, I finished something and I felt really good about it. It's just like, this is my bread and butter. This is my baby. And what happens a lot to me is I get very defensive on basically, like, this is the best ever thing that has been since sliced bread. And you defend it as much as, like, possible. It's like defending a thesis, right? And a lot of the time there's probably something better and newer that you could try out and things like that.

Benjamin Mena [00:46:55]:
There's nothing, never, ever a. This is fully complete. There's nothing that can be done more to improve it. Right? I sometimes get, like, when I was much like maybe like four or five years ago, that was kind of like my mindset. But this company has definitely made me, like, break out of my, like, kind of mindset of like, yeah, we should try that. Like, we should definitely change this up. I don't think, like, this thing that I built over here, I like, it's really good. I really loved it.

Benjamin Mena [00:47:19]:
I spent weeks building this. But the next day it could be like, every recruiter hates this. I'm like, fair, okay, let's figure out something, right? And I think a lot of people have a hard time basically, like, when they, like, get into that mindset to get rid of something that they spent so much time in and start something new. I think, like, that's also like, the biggest problem in AI Like a lot of people don't like change. Right? So yeah, I think that's probably maybe an example of like one of my like failure mindsets changing, if that makes sense. But like, as it comes to failure.

Steven Lu [00:47:55]:
Oh, that's a good one.

Benjamin Mena [00:47:57]:
Yeah, sorry. I like exact examples are hard.

Steven Lu [00:48:02]:
Well, talking about another example, let's like jump into another example question. Like, you know, if you can go back in time, like you've now worked a bunch of years around recruiters, you've been building products around recruiters. If you got the chance to like, take everything that you know now and sit down with yourself in your early days, what advice would you actually tell yourself?

Benjamin Mena [00:48:20]:
Hmm, early days of like building companies.

Steven Lu [00:48:24]:
Let's say like the early days of like maybe intercell or before that.

Benjamin Mena [00:48:28]:
Yeah, yeah. I mean, if you're like an entrepreneur, I think like the best piece of advice that I can say to like my like first year as an entrepreneur would be whoever. Like, you're in the best time of your life, you're in the best learning situation possible. Like, think of it like, worst case scenario, you're like having the most expensive education of your life. But it's like, like skills and value that like, you'll never learn elsewhere. You can't learn it from school, you can't learn it from your employer. You can only learn it from building a company. At least for like maybe like on my like recruitment, like let's say the recruitment side of things and like building products.

Benjamin Mena [00:49:07]:
I was like, I should have told myself back then. So in the interstellar days, I had like this idea, but at the time, again, it was like a data problem and I was really deterred by it. Back in the day, like five years ago, to be like, ah, AI can't do this. Like, I can't get like AI models to do this today. And what I should have told my old self is keep going because like, this is where the world is going to end up. That's if I had a time machine, of course, but because like I had this idea back in Interseller. It wasn't called pin.com at the time, but it was called Northstar, which is basically like pointing you in the right direction of like, who to hire. But I tabled it because it just would take too much time, too much effort, and like, we just weren't there with AI models at the time.

Steven Lu [00:49:57]:
That's crazy. That's awesome. This has been an awesome conversation. You know, I host an AI recruiting summit every year just because I know the industry is changing. I'm getting pulled to do AI talks, which I'm not the AI expert, but I probably just know just a little bit more than a few people. So, like, AI is exciting for me. So this has been just a fun conversation. I think we could probably sit down for another three hours and just go back and forth.

Benjamin Mena [00:50:18]:
In reality, we can do like round two, right? Like, I'm sure, like you're gonna get a ton of comments on your podcast or through social media channels of like, you know what if this happened, it's. Yeah, happy to. We could go through rapid fire questions of those again. I can give you kind of my, like 2 cents on all those questions.

Steven Lu [00:50:38]:
Well, before I let you go, two things like, how can somebody follow you if they want to follow you?

Benjamin Mena [00:50:42]:
Yeah, totally. I think I'm Most present on LinkedIn. So if you go to LinkedIn.com Stephenlu my full name, I was like one of the first users of LinkedIn. So you can follow me there or just come book a demo with us on pin. So www.pin.com right there on the homepage, you can book a demo and see what we're all about.

Steven Lu [00:51:03]:
Awesome. And before I let you go, is there anything else that you want to share with the listeners?

Benjamin Mena [00:51:08]:
Don't fret. AI is going to help you. I swear. I don't know if you want to take that comically or seriously, but I think it goes both ways.

Steven Lu [00:51:18]:
Awesome. Well, like I said, this has been a fun conversation and I feel like I can talk about AI all day in recruiting. I know most people can't, but I'm excited about the future and for those that take advantage of the future. But I mean, here's the thing. We still got to put in the work. As recruiters, our goal is relationships. Our goal is conversations. Our goal is helping our clients.

Steven Lu [00:51:34]:
And my dream is that way we could do more of that. So for everybody out there, make 2025 the year that you absolutely crush it. Make 2025 the year that you chase your dreams and make your goals happen. Talk to you soon.

Benjamin Mena [00:51:46]:
Thanks for listening to this episode of the Elite Recruiter Podcast with Benjamin Mena. If you enjoyed, hit subscribe and leave a rating.