no longer a platform where you can build

AI agents. We are an AI agent. Like

technology should just work. You

shouldn't have to build an agent to get

stuff done. Using Lindy has changed my

relationship with the computer and I I

barely open G anymore. It's kind of

weird. I used to be glued to my inbox

like a real founder. Open is not real

weekend. Like technology is supposed to

be at your service, not the other way

around. I just sat down with the CEO of

one of the biggest AI automation

platforms to talk about OpenClaw, their

new competing product that you can set

up in under 60 seconds and why how we

think about AI automations and agents

just changed forever. So, grab a coffee

and strap in because this is going to be

a fun one. Flo, what is going on, man?

I'm stoked to have you on the podcast.

There's some pretty juicy things that

we're probably going to talk about

today. Anyways, I'm I've used your guys'

platform for years, probably 3, four

years now. I absolutely love it and a

lot of my audience loves it too. So, I'm

super stoked to have you on the podcast.

It's a full circle moment and stoked to

be talking to you today.

>> Yeah, thanks for having me, Brock. And

you have been with us from the very

beginning, haven't you? I'm really

appreciative.

>> Yeah. No, you guys have a great product

and I think one thing just diving into

this. Yeah, my audience is pretty

non-technical and I've always loved how

your platform is so easy to build out

different automations, you know, like

whether it's a meeting prep assistant or

autoleabeling emails. These are all

things that I've used your product for

for years to like help me run my

business. And I think you guys are

making an interesting transition

product-wise and I'm super excited to

talk about today. So yeah, give me kind

of a breakdown like early stages what it

is that you guys are trying to build and

then we can maybe dive a bit more into

like where you guys are evolving to in

the next couple of months or coming

years.

>> Yeah, 100%. So the way we've explained

who we are for the longest time has been

we are the easiest way to build AI

agents for nontechnical people and I

think we had an epiphany of sorts

sometime around last year where we

realized that especially as the

technology evolved and as the models get

better that the easiest way to build an

AI agent becomes almost a contradiction

in terms

>> you know it's almost like like I told

you like hey I'm the easiest way for you

to make yourself a burger at home a

cheeseburger and it's like hey like if

you really want the easiest way to make

yourself a cheeseburger. You're not

going to want to make yourself a

cheeseburger. You're going to go to

McDonald's.

>> Yeah.

>> So, that's kind of is a big big big

mental shift we've made where we are

going from no longer a platform where

you can build AI agents. We are an AI

agent. You don't have to build it. You

can just use it. So, what we just

released last week is called Mindy

Assistant. And the way the way I like to

describe it is like it's an AI work

assistant that helps you get rid of the

[ __ ] So like your email, your

calendar, your meeting scheduling, your

email triaging, your email drafting,

like it saves you one to two hours a day

from the excuse my French, like the

[ __ ]

>> right? No, I love that. And to give a

bit of my like to give a bit of context

to my audience, how are we interfacing

with this now? Is this something where

it's just like off in the back end

running all this automations and stuff?

You don't really see it. Like are you

texting this? Are you, you know, on the

Lindy interface? How's this conversation

going down with our Lindy assistant?

>> Yeah, I think that's a really good

illustration of the the really big

difference between this new product and

and the old one. I'll show my screen

real quick. I think like just showing

easiest. This is the old product and

this is one of the most common use cases

we had with the old product which is

daily digest. So receiving an email or a

message every morning that's like this

is who you're meeting with. This is the

context of your meeting. So what it did

in this case is like hey every morning

you wake up you look at my meetings for

the day and then for every meeting and

for every attendee of every meeting you

research them you look up their LinkedIn

you look up my past meeting notes with

them you look up my past emails with

them and then you bring together this

really big email of like all of this

context

>> right and that was very much like okay

once it does this step move on to the

next in this like it's pretty I don't

know the right word for it maybe like

programmatic of just like very

structured in the way that it operated

right

>> it was putting the AI on the leash M

>> and I [clears throat] think that as the

models have developed the leash has

become unnecessary and in fact has

become harmful because model loses in

flexibility when it's on a leash. The

new product the way it works I have a

bunch of screenshots here. I've had to

compile them ahead of time because it's

uh it's obviously uh potentially

sensitive information but I just receive

this kind of digest every morning and to

answer your question I receive it over

iMessage. So, it's just like texting a

human assistant. And so, every morning

it texts me and it's like, "Hey, morning

flow." Like, "Today's nice day. 61

degrees. You've got three meetings on

your calendar. By the way, you've got

timesensitive email from Brandon. You

might want to take a look at it."

>> Yeah. And I want to cut you off right

here. Or not cut you off, but like add

my add some input here. This texted you.

You didn't have to go and say, "Hey,

Lindy, what's on my email or what's on

my list for my calendar today?" if I'm

correct here. Lindy proactively reached

out to you to text you and say, "Hey,

it's the morning like let me just like

get you up to speed on everything

because like right that's kind of the

process of this now."

>> No, exactly. So, it's just it's it's

very proactive and I think the

proactiveness of it is is a really big

funding principle for this uh for this

product. For example, here I I'll I'll

use another example. It was uh whenever

I receive a timesensitive email, it'll

text me. Or for example, if I'm in a

meeting, Lindy joins the meetings as

well. She's like I use she pronouns. Uh

she's

>> she's taking notes during the meetings

and so forth. If during the meeting I

say something like hey awesome Brock

like let's meet again on Wednesday next

week and you're like sounds good. Lindy

will be will send me a text afterwards.

She's like nice meeting with Brock like

do you want me to send an invite for

Wednesday? Or if I say

>> reading the transcript of the meeting

then using that as context to then be

proactive and kind of create an action

list.

>> That's exactly right. And you know,

after almost every meeting, you have

these like tiny little action items. And

they're like paper cuts. You know, it's

like you need to send a follow-up email.

You need to like let Bob know about the

decision that was just made. You need to

like send a follow-up calendar invite.

And these things like they may not sound

like a lot, but the analogy I've heard

that that I've really liked is like, you

know, I don't know if you know, like 95%

of the airplane crashes happen during

takeoff and landing. So, it doesn't

matter if you're flying from like San

Francisco to Paris, which is like a

11-hour flight. It's the 10 minutes of

takeoff and landing that are by far the

most dangerous. Like, cruising is

actually really easy. And so, if you had

to fly from San Francisco to New York

and land and take off at every city

along the way, it would be a miserable

flight. It'd be very long and it'd be

very dangerous. And so, the analogy here

is like that is what we are asked to do

every day at work. We all want to be

cruising because that's smooth and

that's that's easier and that's less

dangerous. We want to think

strategically about our businesses like

at the 10,000 ft altitude but we are

asked to take off and land 50 times a

day. You want to think strategically but

ah [ __ ] I need to send a follow-up email

to B. Ah there's this calendar event I

forgot about. Ah I need to jump on this.

And so I think the opportunity that's

opened right now because the models have

become so good is like they can run

constantly in the background for you.

There can be your sidekick that's with

you all day that like is in your inbox

with you that is in your meetings with

you and that catches all of those

takeoff and landing moments and that

takes care that takes care of the the

take off and landing for you,

>> right? I love that. That's that's such a

good good analogy cuz it's just trying

to reduce that friction of like all the

manual busy work. And like one thing

I've been using the Lindy assistant

maybe the last like 5 6 days since you

guys released it. And one thing I've

been personally using it a lot for is

kind of just like as a second brain like

at the end of the day I'll literally not

even just like text it. I'll give it a

voice note and say, "Hey," like a voice

memo and be like, "Hey, here's what I'm

working on. Here are the thumbnails I

need to do for this, you know, for my

videos in the pipeline. Can you research

some titles based on my previous YouTube

titles?" And not only will it like get

creative and come up with the titles,

but it will proactively scrape my

YouTube, look at my previous videos,

find the high performing ones, and then

create titles based on that. So, like

for me, it has been the best thing ever

to basically just like give it a brain

dump at the end of the day or maybe in

the morning or even the middle of the

day when I'm stressed and say, "Hey,

what's on my to-do list that I really

need to get done now?" And it will just

go off and like do it for me because

it's connected to all these different

applications that I'm using, which has

just been the most amazing thing about

this.

>> 100%. I do that so much. So, one of my

favorite use cases is at the end of the

week, I message my Lindy and uh I I tell

her I have a message about it somewhere

here, for example. So every week I send

a message to my agent and I'm like hey

my priority for next week is to work on

X right now it's product I do a lot of

product work and I'm like can you please

let me know do perform a calendar audit

for me like look at my calendar next

week and let me know how many hours of

meetings do I have and how many fit into

this priority and then what do you

recommend I shuffle around what meeting

can I cancel what can I delegate what

can I just not attend in order to make

more room for like my actual priorities

also to your point about training it

like this partner that just like compile

this information and dispatches it down

to like different systems and different

notes files and so forth. This is one of

my favorite use cases whenever we have

like whiteboarding like brainstorming

sessions with the team and when we're

done I sent a picture of the the session

to Lindy and I sent a voice memo and I'm

like hey this was a brainstorm about X Y

and Z and it it can see the whiteboard.

So I'm like this was a brainstorm about

X Y and Z. Can you please just arrange

this messy whiteboard into like a clean

note and put it on notion? So yeah, go

ahead. No, I was just going to say

that's so crazy and I want to make a

point because this is it's a perfect

example because a couple of years maybe

like two years ago when I first

discovered like using using your

platform I had that same idea for a use

case. I was like okay I have this like

notepad. I'm going to paste this in and

I'm going to try to figure out an

automation that could automatically like

categorize this for me or you know do

what I needed to do on the back end. And

it was such a pain in the ass trying to

figure out like okay how do I get AI to

analyze the image? How do like what's

the process for that? And then what do I

do here? With the new Lindy assistant,

you don't need to configure all that

yourself. Like a AI is getting better.

So image models are able to extract text

from images to, you know, make that

actually possible now. But like the

whole fact that we don't need to build

the flow now to build this is the best

part about it cuz you just literally

just ask your assistant and it knows how

to build all this stuff.

>> 100%. It should just work.

>> Yeah.

>> Like technology should just work. You

shouldn't have to build an agent to get

stuff done,

>> right? I I 100% agree. So I guess this

brings me to maybe like one of my next

questions. So for the future of you guys

obviously because a lot of you know

platforms say like naden or make.com you

know they're visual interface builders

and so you think of like okay if I need

an automation built out I'm going to go

use Lindy and build it myself. Now that

that is not kind of necessary as much.

Where are you guys trying to position

yourself long term in being kind of like

that go-to AI assistant so to say?

>> Yeah. So the the workflow like Lindy

workflows are still around and you can

still you can still build them and

they're great for like a lot of use

cases for like much more technical

people who have time and and the

technical ability to build those for

certain use cases they're better they're

like more reliable because they're they

have the leash and they're more they're

more cost effective and and so forth but

right now you know this is really like a

return to our roots like we're really

building a Lindy AI assistant which is

where we started initially and I think

the technology was just not ready like

the model wasn't that good enough yet.

Man, when we started it was like before

GP4 it was GPT3.5 4,000 tokens context

windows. Uh it's it's just nothing

worked and even GPT4 at the time blew us

all away obviously but is nothing

compared to the latest frontier models.

That's where we fit. I think perhaps

like a a a comparison that we hear every

so often is uh it is hard in February

2026 not to mention open cloud. It's

just uh it's really been such a

phenomenon. And I think the way I I

compare it is like I compare it to like

Linux versus Mac OS. Like just like

Linux, Open Cloud is this like strange

vibrant explosion coming out of nowhere,

right? Like Linux came out of Linux

tovolt's bedroom. This like weird guy

out of Sweden, right? And then there's

like this huge excitement in the open

source community going on around it. And

and so engineers really sort of crown it

as the new standard. it really blew up

and and now it's interesting because

fast forward I don't know like three or

four four decades three decades Linux

won but in a very roundabout way like

everybody uses at least Unix without

realizing it like your iPhone your

Android devices run on on on Linux your

your Mac your Mac runs on on it's a Unix

operating system and so I think like

this is where I I draw the comparison is

like Linux and OpenCloud are just

wonderful pieces of technology they are

not products and it's not just about the

deployment that's a pain in the ass. The

deployment is a pain in the ass. But you

see a lot of those products that are

like deploy open cloud with one click

that still does not a product make. Once

you've deployed open cloud with one

click and it's still not an opinionated

product that does stuff out of the box

for you. You have to set it up and I

can't tell you how many people I have

heard tell me ah I still have to set up

open cloud like it feels like a chore

you know it's like and what I tell them

is like open cloud is not worth your

weekend man like technology is supposed

to be at your service not the other way

around you know. So most people they

have it on their to-do list. They don't

want to do it. It's a chore. They don't

want to set up Open Global. And I would

say so 90% of people I know want to do

it. They haven't done it. And then like

90% of the people left, they've they've

started and they've actually deployed

it. So like the problem is not the

deployment. It's just they're like ah

okay once it's deployed actually there's

a lot of crap to do now to like actually

get it to be useful.

And so I just think that again it's the

Linux and Mac OS thing like it is not a

finished product and I think that it's

it's an important new building block

that startups like ours might use to

build an end product.

>> Right. Yeah. 100%. I definitely agree.

Like personally I have set up OpenClaw.

I use it and you know I use both Lindy

and OpenClaw uh to be fully transparent.

But one thing I've noticed is I'm using

Lindy for more sensitive information. So

I'm using that I'm obviously texting it

over SMS, but like I'm using Lindy for

my emails. I'm not going to use OpenClaw

personally for that use case because I

just don't personally trust it. You

know, I don't I'm not technical. So I

still have the fear of like, man, did I

set this up improperly? Like am I going

to give it access to something that I

probably shouldn't give access to? Am I

going to give it the wrong prompt? I'm

not am I not going to set up the right

guardrails? Yeah, I'd be interested to

hear kind of where you guys want to

position yourself like as being maybe

that like I don't want to put words into

your mouth, but maybe being that safe

alternative that nontechnical people or

just anybody can use so they don't have

to like deal with the potential negative

downfalls of using something like

openclaw.

>> Yeah, I I think it's hard in like

February 2026 not to mention open cloud.

I think in a nutshell like we're more

secure and we'll turn king like a thous

you know. I think open cloud is is this

awesome thing but look it's it's 80 days

old. You know, we've been working on

this for three years and so there's been

a lot of security flaws coming out of

it. We spent a lot of time on security.

So we have this thing that we call like

an agent shield that protects you and

your data and the agent that builds like

many layers of protection between your

data and your agents such that no one

can prompt inject your Lindy agent. Like

people can't just email uh your Lindy

agent and get it to do stuff like

there's a lot of layers of protection in

the middle and then it's a lot easier to

use like open CLA. I often compare it to

Linux. It's like it's a beautiful piece

of technology. It is not an end product.

It's not about the deployment. You know,

there's a lot of products coming out to

let you deploy Openflow with one click.

But once you deploy it, you still need

to get it to do something useful for

you. It is not an opinionated turnkey

batteriesinccluded product. That's what

we are.

>> That's a good way to put it. Batteries

included product. Yeah, because again

like going back to what I was saying

earlier, I really use Lindy for all the

like sensitive information. I'm not

going to give open call access to my

emails, all that sort of stuff. But

yeah, I'd be interested to hear. So,

have you guys been working on reframing

and recreating this product and

repackaging it into the new Lindy

assistant? Have you guys been working on

this like the last couple of months or

is this something where you guys decided

to like change because you saw what

OpenClaw is doing? Cuz I'd assume this

is something you guys have had in the

pipeline and you guys have been like

envisioning for quite a while now.

>> Oh, 100%. Like we've been working on

this precisely for about 6 months, which

is like an eternity in AI and we really

wanted to take our time to like get it

right. really, you know, even before

these six months, it's building on all

of the infrastructure and the

scaffolding and the technology that

we've built over the last three years.

So, for example, out of the box today,

Lindy can use a web browser to like make

purchases for you on Amazon and like

look up information and all of that

stuff. That is an announcement we made

in nine months ago, which is browser use

and computer use. So, Lindy is able to

use all of those building blocks that

we've been build building over the

years.

>> Right. And that's actually something I

was going to bring up. you. I'm glad you

refreshed my memory. The browser use

because Claude I mean I guess it's top

of mind. Uh Claude Sonnet 4.6 just

released and apparently the browser use

is you know even just a little bit

better than Sonnet 4.5. So, I'd assume

the browser use you guys have is only

going to get better as these LLMs get

better and the AI technology behind it

gets better because that's what really

gets me excited is having an agent that

not only can, you know, book a meeting

for me because it's connected to, you

know, Google Calendar, but it's able to

go off and actually click inside of a

browser, you know, whether that's like

on an actual browser on my computer or a

virtual browser, which I believe is how

you guys do it. What are your thoughts

on the browser use capabilities? Is this

something you guys think is going to

really be the next big thing or is it

going to be more so just like connecting

those apps and only using browser use if

you can't connect to those apps?

>> Yeah, I think it's both. I think like

browser use is going to be like the

ultimate interface for LLMs. But I I

think uh yes, whenever you can use

direct integrations and APIs, uh it's

obviously a lot faster, a lot cheaper, a

lot a lot more reliable. So you should

you should default to that. But you

would be surprised just how few things

you can do with an API that you can do

with a browser. It's interesting like

you know about two years ago scheduling

has always been a big use case for us.

And it's interesting even when someone

sends me a Canonly link I'm kind of lazy

to like click and look at my times and

all of that stuff. I just want I just

want it to happen automatically. And so

we've been in touch with the team over

at Canonly actually just across the

street for a long time. in and we were

like, "Hey, can you guys please give us

an API so if someone sends a Canly link

to a Lindy user, Lindy can book times

through the API?" And Canly was like,

"Not really. We don't want to let you do

that." Like, we don't want to be

disintermediated. We don't want to offer

an API and by the way, it's going to

open butts and all of that stuff. And

and what I told them is, you know, look,

we're going to do this the easy way or

the hard way. [laughter] I like browser

use is coming, you know, and I think at

the time they didn't take me fully

seriously. I think now they do. And uh

yeah, you know, now when I have someone

fast forward to two years, like the

Canly UI is more than simple enough for

browsing agents to be able to handle it.

And so whenever someone sends me a

calendarly link, I don't even bother to

take a look at it. I just forward it to

my Lindy and she looks at my calendar.

She's got my booking preferences and she

just finds a time and books it on the

calendarly.

>> Yeah. No, I love that. I guess that kind

of takes me to my next question. Like

what are some maybe RAM use cases that

you're using the Lindy assistant for

that somebody might overlook?

>> Yeah, this one is is one of my favorite.

It's it's really strange. It's I would

say using Lindy and this is something I

hear from my friends as well and from

our users. Using Lindy has changed my

relationship with the computer. I'll

tell you exactly what I mean. This is a

daily use case for me. I just like

sometimes I'm in backtoback meetings and

I'm emering from meetings after like

five hours. And so I've not looked at my

inbox for five hours. And I usually like

to be very very responsive over email.

And so after these these five hours of

meetings, by the way, I don't even

really feel like looking at my email. I

just feel like going out and take a walk

after five hours of like straight calls.

And so I literally just go out and I

shoot a message to my Lindy and I'm

like, "Yo, is there anything urgent or

important in my inbox that I should take

a look at?" And Lindy responds and she's

like, "Yeah, three things." Boom, boom,

boom, boom. And then you see I I I

respond with a voice memo and it's

magical. It's like I'm talking to my

box. I'm like, "Okay, email number one,

it's actually not a big deal. Just

delete it." Email number two, yeah, tell

him it's okay. We can do this. And email

number three, can you forward it to our

accountant and ask him what to do? Same.

That's over. You know, and so this is

what I mean by my relationship with my

computer has has changed. I use my

computer a lot less. Honestly, I I use

my my mobile a lot more and I I barely

open Gmail anymore. It's kind of weird.

I used to be glued to my inbox like

every other founder. Now everything just

happens on MS through my mobile and it's

kind of weird. I'm looking at the office

right now and I see all of those

computers and those like large displays

and increasingly I view these things

almost as like it's almost like a

there's a little bit of wristfulness.

There's some something almost like

melancholic about it actually where I'm

like I view them as relics of the past.

I I view them as like I view a computer

and a big monitor and a keyboard and a

mouse. I view them as a tool that we use

to micromanage technology, right? You

need a keyboard and a mouse because you

need to micromanage your thing keystroke

by keystroke, click by click by click,

you know, and um and and and this is

you're almost like a factory worker.

You're a knowledge worker, but you're

almost a factory worker. And that's a

machine that you're using to produce

knowledge all day, right? It's a very

complicated machine. And I think that

when your computer becomes intelligent

enough, it's just like an employee like

you don't need to micromanage it

anymore. And so the smarter your

computer is, the more of a high level

code you can get, you can give it

instead of micromanaging it. You can

tell it, hey, go to my inbox, find that

email, put together a report on Google

Docs, convert it to a PDF, send it back

to him. And the all of the gluing

together of all of the apps, all of that

stuff, the computer's going to figure it

out by itself. I think a lot of it too

probably boils down to like how creative

you could get with the tasks that you

want to have it go off and do. So like

that's something for example with Claude

and Lindy that I've been experimenting

with of like okay how far could I take

this thing you know like of course like

I could maybe have it go through my

meetings and schedule a meeting and you

know that sort of stuff but like I've

been increasingly impressed at the stuff

that it can do and like to piggyback off

of what you're saying where it almost

feels like computers are a relic of the

past. I it's funny because we're both

like so immersed into technology and

we're so all about like man for me I

love new Apple products. I love a new

MacBook Pro. I love a new Mac Mac

Studio, Mac Mini, whatever. But like I

grow increasingly more and more just so

tired of looking at my computer and I

would like I I'm so excited for the

point in time when AI is good enough to

where I can give it a voice message, go

on a walk, and it could go off and do

all these different things for me. and

you know, it's doing maybe more work

than I would get done in maybe like 3 or

4 hours of sitting at my computer and

like doing my deep focus work, if you

even want to call it that. So, that's

why I've been so impressed with kind of

this new revolution that we're starting

to see of just these agents that we

could text or even, you know, I'd assume

at some point we could maybe have a

phone call or just like chat over voice

message and have it go and do these

things. That's why I've been really

loving this product that you guys

created because I could reconnect with

nature more. I could reconnect with my

friends instead of having to feel like I

need to grind all day and do these

things manually. It's just like it's

interesting how the dynamics are

changing.

>> 100%. I mean like the way we've always

expressed our mission is to free

humanity from work. Once humanity is

freed from work, leaving aside for now

the deep societal questions that is

asks, but like once humanity is freed

from work and and we figured out those

deep questions, you're not going to have

a computer. You know, you're just going

to be at the park in the sun hanging out

with your friends and family and you're

going to have your phone and maybe

there's a neural link at that point and

you can just talk to your phone. like,

can you bring me a broom? [laughter]

>> Right.

>> Oh, it'd be neat if if there was a

building there. And I I agree with you.

You sort of become bottlenecked by it's

a really surreal feeling because, you

know, we built this app and we ourselves

discover new use cases for it non-stop.

And once you start to make it a habit,

you start small. You start with like the

obvious stuff. Everyone wants help with

their email and and their their meetings

and their calendar. But then little by

little, you use it more and more. You

talk to it all day. And little by

little, you at your computer and you do

something and you feel silly. you're

like, "What? Why am I doing this?"

Probably probably Lindy can do this. And

so, I'll give you an example of

something that happened the other day.

Like, I I'm I'm working a lot on product

and you know, I I I file a lot of

tickets whenever I encounter an issue

with Lindy. And so, I'm talking with

Lindy. Something bad happened. I'm like,

"Ah, this sucks. I take a screenshot on

my phone. I send it to my computer. I

open linear. That's what we use for our

buck tracking. I create a ticket. I set

a priority. I assign it to someone. I

attach the thing. I just And I'm like,

wait a minute. Why am I doing this?"

>> [laughter]

>> And so I just sent this message to Lindy

and I'm like, "Hey, when I ask you to

file a ticket or report a bug, I want

you to just go on in the air and like

file it." And then if if I ask you, you

know, ping someone on Slack about it as

well. So that's what I mean by like

micromanaging the computer, you know,

like here this workflow is sort of split

between linear and and Slack. And so I

think there's almost no workflow that is

contained inside a single application.

It it always spans multiple

applications. you know, you bring

together a slide deck and then you

export it and you send it by email. And

you know, today the workflow has to be

orchestrated across all of those

applications. And the guy who's got to

do the orchestrating is the user. And

and no one thought to themselves as a

kid, when I grow up, I want to be a

workflow orchestrator. And and so that's

my point about the relationship with the

computers that's changing where you can

just give a goal to your computer and

it's going to orchestrate the workflow

for you.

>> Yeah, I think I 100% agree. I think

that's just a great way, you know, a new

direction we're moving in. And then I

guess like one question I have is how is

Lindy learning like how you operate for

example because I noticed that when I

was inside of my Lindy assistant that

there was this like I think it was

instructions part of it where I noticed

it was like my email agent and it

actually read through my emails,

understood my my language and how I

actually respond to specific emails. So,

I'd love to hear because obviously these

AI assistants are kind of bound to like

how good of a responses they give, you

know, but if they don't understand how I

actually respond, then what's the point

of having it go off and do it for me?

Nobody likes AI slop. But I want to know

your process for that, like how is Lindy

actually learning, you know, how we

speak, how we operate, and how we

actually perform different things.

>> Yeah. So this has really been like a

journey internally where like we spent a

lot of time developing this system so

that Lindy gets continuously better and

like learns your preferences and and it

becomes magical once it really works.

Like there is this example where Lindy

sent me this email. She's like hey your

lawyer just sent a counter offer to

these guys. And by the way I noticed in

the counter offer he uses this address

for your office. Is that the right

address? I thought this was your office

address. And I'm like oh yeah no you're

right. Like he got the address wrong.

Like please send him an email.

>> Hey that's crazy. That's actually

amazing.

>> And I I'm not lying when I say like

every day, including like a couple of

hours ago, I was meeting with an old

friend and Lindy sends me always a

message when I meet with someone with

like a prep for the meeting and the

agenda who I'm meeting with. And uh she

sends me this text message and she's

like, "Oh, you're meeting with Shukont

today. He's an old friend. You've worked

together for five years. He was like an

OG on the crew a while ago." And I'm

like, "How the [ __ ] did you [laughter]

know that? How did that how did that

happen?" And so what happened is like we

built this system. this like continuous

learning memory system. We iterated on

it for a long time. We really made it

shine like a beautiful jewel and and

we're like, "Oh my god, it works. It's

awesome. Little by little, it's learning

about me and it's it surprises me in how

much it knows about me." The only

downside is like you really only reap

these benefits after you've talked for a

while about about yourself, right? It's

a relationship you build over the long

term. And so I remember distinctly we

were at the office late at night when we

were like jamming about this, having

dinner with the team, and we were like,

"Wouldn't it be neat if Lindy could just

know about you from the get-go?" like

how might we get all of that data about

you, all of these preferences. And so

what we ended up doing is we we built we

called this the hydrator system. So it

hydrates your knowledge base and your

memory base. And so what we do is when

you sign up uh we we ask for access to

your inbox, we again suck to compliant,

HIPPA compliant, super safe. And if you

give us access to your inbox with your

permission, we look at your history of

emails. We look at thousands and

thousands of emails and we process them

using LLMs. Again, nothing is ever used

for training. We have agreements in

place with the labs and so forth, but we

we process all of that and we use it to

learn about you. And then it's true, you

can go to your settings and you can see

everything we've learned. And very often

people find it delightful actually to

learn to like read this report and we

use it then to do a lot of stuff

including drafting replies for you

because we've learned what you reply to

whom and and how you write. And I'm not

lying that at this point it's a daily

occurrence. You know, I I look at my

inbox and I look at a draft that's

sitting in my inbox to an email I

received. And at first I thought that

was I was going crazy. At first I

thought I was like, did I write this

draft? I have no recollection of writing

this. And it's just like it's Lindy who

wrote it. So yeah, that we've paid a lot

of attention to that. Like not just the

fact that it's proactive, but the fact

that it's personalized and keeps

learning about you and getting better.

>> Yeah, that's that's huge. And like you

said, the fact that you don't, you know,

of course it could learn the more and

more you chat with it, but the fact that

it already has that information just

because it's able to scrape it. Like for

me, I probably wouldn't have allowed it

to draft email responses for me if I if

it didn't already have that pre-existing

knowledge, but now it's just so good.

I'm like, "Okay, this is good enough to

actually use." So that's been, you know,

something um amazing for me when I've

been using it. Is there anything else

that you'd want to chat about or any way

that we could kind of end it?

>> It's available today. [laughter]

Yeah, absolutely.

>> Um and um no, you know, we all we are

hiring. If anyone is listening that's an

engineer who's excited about this kind

of stuff, we we are growing the team

rapidly.

>> Perfect. And then as for pricing, what's

what's pricing look like right now?

>> Yeah. Uh it cost $49 a month. Got it.

>> Huh?

>> No, I was going to say and how far does

that get you? Because I've been using

it. It's gotten me very far. I haven't

hit any limits. So I'd assume for the

majority of people that is probably

sufficient unless you're using tons and

tons of usage of this.

>> Exactly. It's enough for 98% of people.

There's like 2% of people who run into

limits and we just tell them like hit us

up and and usually what we do is like it

becomes like $100 a month, sometimes

$200 a month for like the very very

heavy users. Some people use for coding

which like never really intended but

like hey we can't really give you

infinity tokens for $50 a month and so

yeah it's enough for 98% of people.

>> Got it. Yeah 100%. Well, Flo, thank you

so much for coming on, man. It's again,

like I said, full circle. I've been

using your guys' product for years now.

Super stoked to chat about this and I'm

excited to see what you guys keep

building and hopefully I could have you

on again when you guys have another

groundbreaking kind of product release.

>> Be a pleasure. Yeah. Thank you so much

for having me, Brock. Cheers.