– English Lyrics

Lyrics & Translation

[English]

[Music]

Today I'm joined by Eron Wong and John

Wright, the founders of Alex. They just

announced a 17 million series A with Pak

15. Congrats. That's awesome.

>> Thank you. Great to be here.

>> Can you first tell us what Alex does?

>> Alex is your AI recruiting partner that

helps companies and staffing firms

interview everyone and ultimately hire

the best people. Alex has access to all

of your favorite recruiting tools and

can conduct phone screens, video

interviews, scheduling, sourcing, uh,

updating the applicant tracking system

completely autonomously.

>> So, what's wrong with our today? Like,

what are you trying to fix?

>> Sure. So, applicant volume has tripled

in the past 3 years and time to hire is

at an all-time high. It's about 60 days.

I mean, we're here sitting in San

Francisco. You can close a house in San

Francisco faster than you can hire your

next engineer. And so the hiring market

is incredibly inefficient today and it's

never been more efficient in history.

And the main bottleneck is uh one

bandwidth of the recruiters uh and two

uh matching right how can I actually

match the best candidates with the best

opportunities. So, Alex solves both of

those problems by allowing any company

or any employer to interview everybody

at scale, right, and get incredible

insights on these candidates and to be

able to match those candidates to the

right opportunities because we have

access to all that data.

>> Can you give us a sense of like the

scale uh of the operations here? Like,

how many interviews have you performed

so far?

>> We're doing thousands of interviews

across all of our customers every single

day just since we launched last year.

Uh, Alex has helped hire thousands of

people and today Alex is uh filling tens

of thousands of active job roles.

>> Who are you doing that for? Like is it

mostly mostly like companies or staffing

agencies or how does that work?

>> What's really interesting about kind of

the working in the recruiting space is

that uh it's an incredibly large market.

Every employer hires by definition,

right? And so the question is where is

that niche edge of the wedge? Where do

you get started? We found a lot of pull

from the staffing agencies. Our

incentives are highly aligned. If we do

a really great job

>> is that you're more people. Is that the

main

>> Exactly. Right. They concentrate the

pain point. So it's a true hair on fire

problem. If we do a great job, we also

increase revenue for them. That's how

they make money. And so uh staffing is

one of our largest segments. We do also

work with large some of the largest

employers in the world. Um for example,

the largest US supplier of let's say

nucle like nuclear components, right? We

have a huge electron deficiency with

training.

>> But what roles are you hiring for here?

>> Yeah, it's across the board. Everything

from highly technical uh staff software

engineer roles to senior accountants to

for that nuclear company. They're a

publicly traded nuclear company. We help

them hire nuclear welders, right? And

that's much more blue collar.

>> And so the AI do you test the technical

skills too? It's like stays kind of like

first screen kind of interview. Yeah,

that's exactly right. So, it tests for

skills on that first call or that first

interview. What's really interesting is

nuclear welding is very niche uh

discipline, right? But with uh Alex,

right, with an AI, you're able to really

dig deep into that discipline and really

better understand if that particular

candidate is qualified for for that

role.

>> And how do you make sure the

interviewers the interview the AI

interviewer knows about welding?

>> Yeah. So, um, first, Alex has access to

your applicant tracking system or your

HIS. And so, uh, Alex will be able to

see the types of people that you've

hired before. We'll be able to see the

job description and any intake meeting

notes from the hiring manager. And so,

Alex will learn to be the best uh,

employer brand ambassador for your

company. Alex uh can also be fine-tuned

on particular roles like nuclear welding

um such that uh it will find and and and

ultimately qualify the best people for

the job.

>> Let's go back to the early days of Alex.

You're both relatively young, right? I'm

not sure you did a lot of interviews

yourself.

How come you picked that idea?

>> We've been candidates a long longer than

we've been employers actually. Um but I

think that ends up being a really unique

advantage. I think for a lot of uh the

companies that have tried to build in

the recruiting tech space today, I think

where they haven't succeeded is in

building uh a candidate experience uh

that you know really puts job seekers

for first an experience where hey I

applied for this job I want to talk to

you know an AI want to engage with this

employer right and so I think that gives

us a very unique advantage you know John

and I met in undergrad we met at Brown

um we had built a hiring uh tech company

uh before this um and that was a lot of

fun.

>> Learned a lot and yeah um it really kind

of helped us better understand the

market what like the pain points were

for uh for candidates and optimize for

the candidate experience cuz ultimately

um that's kind of what's more most

important uh in what we're building.

>> So how long have you been working on X?

I mean like a years ago like you

couldn't even imagine using AI for that

right? I guess now now you can make that

conversational you can have a real

conversation with an AI. When did that

become possible?

>> It really got started uh in uh the end

of uh 2023 with the launch of G GPT4

Turbo. We ultimately we're building at

the start of 2024, launched towards the

end of the batch. I believe it was in

April of 2024 um and quickly got our

first customer and then uh continued to

to scale from there.

>> What were the biggest challenges? You

know, like there is that uncanny valley

where it's so weird to speak with an AI.

It looks like we're past that. How did

you get there? Yeah, at the end of 2023

is when this kind of building a voice

agent that was had low enough latency um

where candidates were comfortable

talking with it. Um that's when that

really became possible. Um and there

were no other voice tools out there to

kind of build this like you have today,

the Vappies, the retails. So we really

owned that orchestration platform and

being able to connect all the different

models succinctly to be able to have a

low latency, high performant um reliable

agent. And over time models just kept on

getting better and better. So we made,

you know, incremental progress on

latency, on voice quality, transcription

quality, and that really helped kind of

optimize the candidate experience.

That's actually one reason we named the

AI recruiting partner Alex was because

the transcription models weren't quite

there yet in uh in even in 2024. And so

like Alex was like a it's a very easy

name to understand that the trans that

the transcription models could like

understand is say like Alex okay like

that's very obvious that he's here he

>> text would always get his right

>> exactly right it would always get it

right um and so that's kind of the

origin of the the name

>> the conversation during the interview

can lead the in some kind of like

corners or corner cases but at the end

you want to have some score how do you

make sure you you get back get the

conversation back on track when

necessary.

>> That's just a matter of prompting um and

a lot of kind of um really seeing these

candidates do do these interviews,

evaluating your agents at scale is just

making sure that their guardrails in

place so that kind of um the questions

all questions are asked, follow-ups are

asked appropriately and making sure that

um the yeah, candidate has no opening

for for trying to to take advantage of

of Alex, even though we've had a lot of

candidates try doing that. No successful

attempts.

>> Do you have any good stories of

candidates trying to hack the AI?

>> We've definitely had uh interesting

attempts, you know, candidates,

especially software engineers who are

kind of trying to break and see what

what the um weak points are of of the

agent. Um we've seen, you know, people

try to do that and usually at the very

end we, you know, we ask for, hey, do

you have any feedback? Like what's the

interview? How'd the interview go? And

they're like, yeah, I'm I'm impressed. I

wasn't able to break it. Um so we we've

definitely had a lot of Yeah. people who

applied to for our software engineering

roles to try and

>> so did you score them higher in the

>> we've had candidates like try to talk in

like XML or in markdown to try to prompt

inject unfortunately for them that that

hasn't worked um but it's it's funny an

interesting phenomenon here is that um

you know when you build you know agents

in a particular vertical for this one

where you're in the job market it's a

market right and so one clear dynamic of

a market is that you have uh adversarial

effects right that is you know

candidates are going to do what they can

to try to get the job right and you know

the job is a scarce resource and so

recently we've seen a lot of uptick in

candidates using you know their own AI

to uh mass apply to jobs but also during

the interview have you know transparent

you know overlay like a clue or

something like this to try to cheat on

uh interviews and so that's something

that we've had to build out building

kind of cheat detection to make sure hey

this person is legit right we've had

we've had to models for detecting deep

fakes, right? Because now they're like

live deep fakes that you can stream into

your device.

>> Yeah. Because you see the video of the

candidate or

>> Exactly. Right. Because these interviews

are often times done through like a

video conferencing platform that we

actually built from the ground up.

>> And so like interviews that end up like

being an a candidate agent and then the

recruiter agent and they speak to each

other.

>> Yeah, that's that's it's it's funny. We

get that a lot and we'll see where the

future goes, right? Like that's actually

pretty interesting. um that the the two

sides of the market, the employers and

and job seekers will will always

challenge each other. But ultimately, I

think it's it's it's actually good,

right? Because that means that the

market's evolving, right? You know, we

believe in a world where, you know, jobs

should be much more accessible, right?

Today, you know, we see knowledge is

democratized, right? I can go on Google,

I can go on Chad GBT, education is is uh

a lot of education is democratized,

right? But still like the job market

opportunities are not right. I'll apply

to hundreds of jobs as a candidate and

I'll be lucky if I get a handful of

interviews. We believe in a world where

if you apply for a job, your dream job,

you should be able to get that

interview, right? And it shouldn't be

about if you, you know, went to an Ivy

League or, you know, whatnot. It should

just be, hey, can you tell me about the

skills you bring to the table and the

knowledge that you

at least get that first interview?

every candidate gets a

>> nurse like some overlooked candidates

that would not have had an interview.

>> Yeah.

>> Like do you have like any examples where

you actually ended up helping a company

hiring a candidate they would never have

interviewed?

>> Yeah, there are several examples. I was

actually just on the phone with one of

our customers yesterday. Um so Cobalt is

this very old programming language. Um

it's used by some of the largest banks

and financial institutions in the world.

I think it was started in 1959. not a

lot of Cobalt developers left in the

world, right? And so one of our

customers is the staffing firm that

hires for all those Cobalt developers.

When they launched Alex, um Alex found

um looked through their database of

their own app through their own

applicant tracking system, found people

that seemed relevant for this cobalt

developer role that they were hiring

for, automatically reached out to them,

interviewed them, and found 11 people

that they automatically just submitted,

and they got placed to their clients

right away.

>> And they were lost in the

>> they were lost in the database, right?

Whether that database is your own

applicant tracking system or your own

applicant pool or sourced from LinkedIn

or Indeed, you know, these pools are

just getting larger and larger and

larger and much more harder to handle

with, you know, manually. And so our our

aim is to really be that recruiting

partner that says, "Hey, well, you know,

there is that diamond in the rough.

Let's use AI to help you find it."

>> Now, you might have already talked to

the perfect candidate, you know, years

ago, and you know, with the current

technology, it that person is just lost.

But with Alex, you know, you're able to

create opportunities that just

previously, you know, you wouldn't have

been able to.

>> Yeah. Is there any specific metric

you're tracking closely like kind of any

evidence for example that really working

for for your customers?

>> Yeah, there are several. Um, I think one

of them that we typically see in pilots

is can Alex given that they interview

all these folks,

>> Alex stack ranks them, will Alex's stack

rank be similar or even better than the

stack rank that uh a traditional

recruiter would look at? Right? So, you

know, is that best candidate really

better than the second best candidate

that's really better than the third best

candidate, right? And I can say that

because, you know, I have the video

interview. I can actually pull quotes

from that interview and say, "Hey, yeah,

this person scored a 95 on their uh

let's say their Python knowledge

because, you know, they have, you know,

10 years of experience. They said this,

they did that." And I tried to dig

deeper and they, you know, kept pulling

out evidence from their experience and

this person's fantastic. And so that's

one way we look at it. Other things that

we're seeing um are increases in

retention rate, right? So these

candidates that are getting hired are

actually staying longer at their jobs.

And also for a lot of our staffing firm

customers, again, they generate revenue,

right, from placing people. And so if

they the client hires them and then they

don't end up staying, they don't they

don't get any incremental revenue. And

so that's another way that we track

success for for our customers. And

again, it really ties back to actually

the candidates. So the first thing we

really look at is our candidates

enjoying, you know, interviewing with

Alex, meeting Alex.

>> How do they feel? How could they feel

like like I guess the first time was and

still there's a lot of first times

today, right? A lot of people have never

spoken to any.

>> How do they react?

>> I think my favorite part that I hear

from candidates is that we really

eliminate ghosting, right? Because in a

world where, hey, if you apply to this

job, you get that opportunity to at

least be heard, right? You get to be

updated on your your your candidacy,

right?

>> Like versus the company where they never

hear back. At least they get the first

interview even if it's an AI.

>> Exactly. And they can also ask questions

during the interview. They have Alex's

phone number and so they can text Alex

at any time and say, "Hey, you know,

actually I had this question about your

PTO policy or hey, are there any updates

to my application? Has this application

been filled yet?"

>> It's more that interview. It's the whole

candidate experience.

>> That's exactly right. Yeah. but really

supporting the candidate throughout

their entire journey at any particular

recruiting or hiring process that that

they're in.

>> Is there anything that Alex actually is

looking for in candidates that u human

interviewers are overlooking?

>> There's an incredible amount of data in

within a video interview, right? And a

lot of it is is missed, right? If you

think about a video interview, 20 30

minutes, you know, that could be, you

know, gigabytes worth of data. You know,

with a traditional recruiter, that's

oftentimes boiled down to handwritten

notes, right? Um, but with with Alex,

that's all recorded. You know, you have

much more in-depth notes. Um, and Alex

is able to both analyze, hey, your

technical skills, your hard skills,

requirements for the role, but also

things like soft skills, right? If I'm

hiring for, let's say, a salesperson, I

want to make sure that they are a

concise seller and they have concise

communication, right? That's something

Alex able to test for because Alex has

that that video data, that audio data,

and will remember that that interview.

>> Okay. Awesome. Let's talk about the news

you are announcing. So 17 million siz uh

that you you just closed. Uh where will

the capital go? Why raise now? Like tell

us more about the the round.

>> Now is an incredibly exciting time I

think to to build um a company in

particular. Um but we've seen incredible

market pull across all of our customers.

A year ago I think we wouldn't have been

able to say that you know especially you

know hire selling to some of the largest

companies in the world. you know, you

know, HR is one of the areas that is,

you know, you know, they've really been

burned by technology that hasn't worked

for them in in the past. And so, a lot

of what we've been doing recently has

been building that transparency,

building that confidence and building

that relationship with with some of the

largest employers in the world. And

that's really allowed us to um one uh

sell into these organizations and two

continue to build a great product for

them. And so we're going to use this $17

million to really make sure that hey, we

have the best product when it comes to

uh AI and recruiting today.

>> Uh and we want to make sure that we

continue to be best-in-class and

supporting uh our customers in in that

way.

>> All right. So just having a better team,

>> growing the team.

>> That's right.

>> Okay. Anything non obvious?

>> Anything not obvious? You know, I think

a lot of it actually is relatively

obvious. I think something that is less

obvious is being in San Francisco,

you're around a lot of companies that uh

hire a lot of engineering and technical

talent and will oftent times like

outweigh the goto market side of things.

I think for us um go to market is

extremely important. You know, things

like marketing, design end up being

incredibly important because again, you

know, of course you need to have a great

product, right? You need to create value

for your customers, right? You need to

make something people want. But in

addition to that, when you have

enterprise sales, right, relationships

also matter, right? Building a brand

also matters. Go to market at the end of

the day matters a lot. We'll likely see

kind of more go to market hiring than

probably the traditional uh Silicon

Valley company.

>> You're also announcing a a rebranding,

right? uh your name was not Alex. Alex

was the name of the interviewer. Like

was that creating some confusion or why

did you choose to go from a prior to

Alex?

>> Yeah, Alex was uh the name of the an AI

recruiting partner and a lot of our

customers were already calling us Alex.

Um and so it made a lot of sense to

reduce any type of confusion and just

rename uh the company to to Alex. Um we

had already seen companies do a great

job with this anyways. You see saw this

with a kodium right turning into wind

surf and the particular naming the

company actual name made a lot of sense.

You're seeing a lot of success with that

with like Harvey for instance. We want

Alex we want the company we want AI to

be approachable and and grounded right

it's not just another piece of abstract

AI software. It's really you know your

partner.

>> Have you seen people react differently

to that that new name?

>> Yeah. Well, they can certainly, you

know, pronounce it a lot easier uh and

share it a lot easier, but uh we've

we've had nothing but, you know, great

things or heard nothing but great things

from our from our customers and uh so

far. And so we're we're really excited

about the name change.

>> Looking ahead, like what do you think

will still be uh I don't know, five

years from now, will there still be

human interviews? Like where are we

going?

>> We're not looking to replace recruiters.

Um and recruiters won't be replaced.

They're going to be supercharged, right?

a lot of their time today when it comes

to interviewing is you look you're

asking the same you know five questions

to this these you know hundreds of

people there's a lot of scheduling

you're updating your pieces of software

right adding notes into your applicant

tracking system um and we want to give

that time back right a lot of those

administrative tasks can be uh pulled

out and you can reinvest that time into

more strategic roles for instance uh

building a relationship with your hiring

manager helping close that candidate

that's on the fence right or spending

more time with the candidates that you

know are qualified and you really want

them to join the team right those are

things that I don't think AI will be

able to replace right those human

elements and good they shouldn't but

those more administrative tasks we

certainly want to be able to pull them

out of that and and help them do more

>> and speaking of hiring so you are

expanding the team like any role you

want to uh to pitch to to the audience

>> yeah uh we're hiring I mean on every

front so engineers um both full stack

front end backend designers uh PM M um

go to launch to go go to market as well.

>> Yeah. If you want to join a company

that's not replacing people but helping

hire them, right? I think that's like

very exciting and that's the future that

we believe with AI and it's the future

that we want to build towards.

>> That's awesome. Hey, before to conclude,

is there anything you you wish you knew

when you started looking back in time

two years ago?

>> You know, we get this question a lot,

especially when we come back to YC and

and talk with the current badges. things

that stick out to me are I think at some

point you need to you know put your foot

down and have really high conviction in

what you're building. there's a delicate

balance between you know listening to

the market and and and make something

people want and saying hey look we know

that this is going to exist in five

years question is like in what form we

came out of the batch we launched um we

knew that there was going to be some use

case for an AI recruiting partner we

weren't sure if that was going to start

with employers with tech companies with

you know the you know fortune 100s you

know there are a few of our customers

today with staffing agencies and so I

think you need to be really obstinate

about that vision that you have but kind

of flexible on on actually how you're

going to get there

>> as some usual advice for founders from

your journey.

>> Yeah, I mean you just need to be

extremely versatile and be able to grow

uh kind of we live in an age where

technology is changing and maturing so

quickly and you you know there's new

models, new new use cases coming out.

So, just being able to really grow with

the technology and say, yeah, just kind

of have a clear long long-term vision

goal, but be flexible on how you achieve

it and kind of what um what tools you

use. You can really learn anything in a

very short amount of time. So,

possibilities are endless and you just

need to be super super um excited to

build.

>> Eron John, thank you so much for joining

us today. It was awesome to catch up.

So, thank you.

[Music]

Key Vocabulary

Vocabulary	Meanings
recruiting /rɪˈkruːtɪŋ/ B1	noun - the process of finding and hiring employees
interview /ˈɪntərˌvjuː/ A2	noun - a formal meeting to evaluate a candidate for a job verb - to question someone to evaluate their suitability for a job
candidate /ˈkændɪdeɪt/ A2	noun - a person who applies for a job or is being considered for a position
autonomously /ɔːˈtɒnəməsli/ C1	adverb - acting or done without external control or influence
inefficient /ɪnɪˈfɪʃənt/ B2	adjective - not achieving maximum productivity with minimum wasted effort or expense
scale /skeɪl/ B1	verb - to increase in size, amount, or degree
niche /niːʃ/ B2	noun - a specialized segment of the market
optimize /ˈɒptɪmaɪz/ B2	verb - to make as perfect or effective as possible
transcription /trænˈskrɪpʃən/ C1	noun - the act of writing down something spoken
retention /rɪˈtenʃən/ B2	noun - the act of keeping or continuing to have something
strategic /strəˈtiːdʒɪk/ B2	adjective - relating to the identification of long-term goals and the resources needed to achieve them
versatile /ˈvɜːrsətaɪl/ B2	adjective - able to adapt or be used in many different situations
conviction /kənˈvɪkʃən/ B2	noun - a firmly held belief or opinion
mature /məˈtʊər/ B1	adjective - fully developed or grown verb - to become fully developed or mature
obstinate /ˈɒbstɪnət/ C1	adjective - stubbornly refusing to change one's opinion or course of action