Hang tight while we fetch the video data and transcripts. This only takes a moment.
Connecting to YouTube player…
Fetching transcript data…
We’ll display the transcript, summary, and all view options as soon as everything loads.
Next steps
Loading transcript tools…
This AI Marketing Tool Automates Everything | Ayra Founder Mimmi Liljegren Explains (Full Demo) | The Builder's Mind | YouTubeToText
YouTube Transcript: This AI Marketing Tool Automates Everything | Ayra Founder Mimmi Liljegren Explains (Full Demo)
Skip watching entire videos - get the full transcript, search for keywords, and copy with one click.
Share:
Video Transcript
Video Summary
Summary
Core Theme
This podcast episode features Mimi, CEO and co-founder of IRA, an AI agent that automates marketing and communication tasks. The discussion covers Mimi's entrepreneurial journey, the development and functionality of IRA, and advice for aspiring founders.
Mind Map
Click to expand
Click to explore the full interactive mind map • Zoom, pan, and navigate
Hi everyone, welcome to Builder's Mind
podcast. In this episode, we're going to
have Mimi. She's the CEO and co-founder
of IRA, an AI agent which automates all
your marketing and communication task
for you. It does the research, it does
generate the content and also it
distributes and then it gives you a nice
dashboard of what's happening with your
SEO, with your marketing, with your
posts. So we're also going to talk about
Mimi's journey into startups, how she
came up with the idea of IRA, a live
demo of IRA to see how you can automate
your marketing and communication tasks.
And also we're going to chat about the
go to market playbook of IRA and how she
thinks you can sell to companies. And
we're going to ask her some interesting
questions about her startup journey,
what she does on the weekends and also
some advices for the first-time
founders. So let's enjoy the episode.
Hello Mimi, welcome to builder's mind podcast.
podcast.
>> Hi, thank you so much for having me.
>> Thanks for joining. Looking at your
journey, it seems that you have quite an
extensive knowledge and experience in
communication part. So I would love to
hear that what happened you know that
you decided to actually build a startup
for this space and what was that exact moment?
moment?
>> Yeah. So uh the moment was about a year
and a half ago. Uh so I've been working
with marketing and communications uh in
an agency uh for the last eight years
and I've been uh holding lectures in AI
for the last three years and a lot of
workshops and about a year and a half
ago I started you know see the increase
in the quality of the content as so I've
been working at the content agencies you
know the quality of the content has been
a central part for so long and it was a
heavy discussion the last couple of
years but about a year and a half ago I
saw you know the potential in the
quality that they could match the human
uh like capabilities in production
definitely in text we weren't really
there in images h especially not when
we're talking about automating image
generation by AI but then uh yeah about
a year and a half ago I saw the
potential and I realized that exactly
what we're doing on the agency side and
also on the in-house marketing
communications team I realized that we
could do through a technology instead of
through people h with that said not that
I mean that we should take away all the
people from this uh you know it's from
this birth but we can definitely
increase the efficiency so much so that
was like the starting point of the
vision of IRA and yeah that was about a
year and a half ago and it's about a
year ago that I launched the company
>> yeah uh kind of like before recording
like you shared a bit of a story kind of
how did you start IRA you know with
technical co-ounder so if you can you
know walk us a bit kind of like towards
that what was your first step for
validating the idea and I think that
that you already knew this space so like
you kind of validated by your experience
but how did you you know take the idea
and then move forward and then start
building and bringing the team together.
>> Yeah. Okay. So I started about a year
and a half ago I started building AI
agents myself. So I'm not I don't have
any kind of technical background but I
realized a lot of that knowledge was
actually shared online. So I realized
that I could get that knowledge from
other sources you know YouTube X
different kind of social media platforms
communities. So I started just like
digging deep into that and I started
building agents myself and at that point
I did it for the agency that I worked on
and their clients and I realized you
know this is actually working out very
well and that's where I started like
thinking could this be a tool for
marketing communications team to work
with and then I started with like you
know the front end how could I do this
so it was it like platform that users
could log into and use on a daily basis
And um yeah, it was uh it was a journey
uh to do this without any kind of
technical background but it was lots of
fun. I learned so much like more than I
probably have done in the last couple of
years. Um and I started you know taking
help of freelancers as soon as I got the
first clients on. I definitely helped a
lot of freelancers helped me out to you
know keep building on the parts that I
couldn't and then I reached a definitely
a point about eight nine months ago when
I realized like I need that part person
as a partner and I've been looking for
that partner since the first idea about
a year and a half but then you know it
takes time to find that partner and I
let that save time I didn't want to rush
into it because I had a vision and I
knew over time I'm going to find that
person so I found Lars, my co- co- um
founder uh that started joining me
part-time in January and then before
summer full-time. H so it's been a
journey and I'm so happy to have him on
board because I'm happy not to have the
technical pressure on me anymore. He's
facing that on the daily basis. uh but
definitely helped me you know also to be
a good co-founder and a good CEO of a
technical company to have the respect
and also the understanding of what it
takes to build products and and you know
the faces to actually have that quality
that we need for for our clients. So L
is saying saying that I hope it's not
just being kind. I hope he's saying it
because it actually means something. But
he says that that is helping out that
I've been trying to build this on my
own. So I have a lot of respect for his
work too.
>> That's super nice. Yeah. Kind of having
the right person next to you building
the company helps a lot. But also that's
what makes the difference you know
otherwise it's hard to build a company
um alone. Uh because it's a like lonely
journey itself and if you're alone it's
more lonely. Uh so I mean I would love
to have you know a short pitch of IRA
what it is for the users and what it
does exactly like some value
propositions as well.
>> Yeah. Okay. So what we do is uh pretty
much we build a branded AI agents that
automate their marketing and
communications. Most common is that we
work from you know starting phase in
marketing communications to the end.
That starts with like medium marketing
intelligence, researching, planning and
then the creativeness, you know, coming
up with creative ideas and then the
productions, you know, producing text
and um images. Uh soon coming video
that's not released yet. And then
distribution, definitely like sending
that material out to all the end points
and then uh doing an analysis of it. So
we're working full scope here. Uh so we
could be like a plug-in product that
works for a whole marketing team even if
you're just like one person or if you're
50 people. H that what we do and the
value that that our clients gets from it
is you know number one when it comes to
AI time we have about 95%age of time
savings in these passable you know
marketing communication tasks. uh but
also quality uh so this is a major part
for us when it comes to content like
it's actually quality improvement when
you work with AI agents and that have
different kind of processes where you
can like proofread stuff over and over
again with like objective perspectives
and also get more tailored content for
target groups
but then also I would say the increase
in production so most the or the average
amount of content that that our clients
produce through the AI agents has
increase of about 4 to 500 percentage
when they've been in the platform for 3
months and there's a massive
increasement. They get so much more
content out there both organically but
also paid material. So those are some of
the you know greatest advantages with
working with AI agents.
>> Yeah, that's super nice. Actually I feel
that marketing part is a bit clear but
then terms of communication. So is it
more like like you guys work with the PR
and then kind of like with media or know
what communication part means here for
the audience? >> Yeah.
>> Yeah.
>> Yeah. So we tailor depending on what
client has depending on their needs. So
it can be you know internal
communications. We have some clients
that always like record the big internal
meetings. They we we package it into
like an internal letter. We send it out
for or through their internal systems.
But it can also be definitely more PR
like picking up different kind of
perspectives, picking up like when
you're mentioning media, but also
preparing press releases and having that
push to different kind of target groups
or end points. Um so we actually
tailored quite a bit. So today when I'm
going to talk about it, it's going to be
more of a set platform with set um kind
of um you know things that agents will
do. But in the real case, I would say as
soon as we work with bigger uh
departments, it's always a a process for
us to adjust and build custommade
processes and agents for them.
>> Perfect. Let's jump in the live demo and
see what it is and how we can do it.
Okay,
so here we go. Uh so
at this um uh demo we're going to I'm
going to do live demo as if you were a
company or working for a company that's
called a bike company. Uh it's
definitely not a real company. It's made
up. Uh but you know we learned this uh
AI agent working in the background
everything about bike company. So all
the kind of target groups, what kind of
reference material, you know, their
channel strategies, what kind of
marketing communication strategies that
you have and also the context is working
in. So we we built up those agents
working behind this interface.
And the first thing the AI agents
typically do is of um it's kind of like
research intelligence. We call it media
market intelligence. And what is
actually that it's combines different
kind of techniques that we already have
out there. So for example, a lot of PR
uh departments actually have different
kind of uh tools that will notify the
department as soon as you are mentioned
in press or PR. So for B bike company uh
let's say they being me mentioned in a
article from bicycling.com
uh then in in the live states uh that
that news will uh pop up and they will
get that article straight in here.
So it combines that kind of mention when
you get mentioned in press or PR you get
that article straight away. You can also
see whereabout your reference and you
can also go and visit the URL.
Um, so mentions is one part and then the
other part is actually what's up and
coming and been like out there for the
last couple of years is how AI can
scrape social media and find you know
trending topics to see what is trending
within your industry and within your
different kind of services or products
that's out there right now. So we can
also do that. So social media trending
is a hu huge part for all companies that
work more you know B2C more social media
strategies and stuff like that.
And then we have the third part that is
news related to to the company. So you
know regular kind of search queries the
search phrases. So we have AI agents to
pull in those search queries those kind
of search phrases and get relevant news
directly directed to the company. Uh so
bike safety tips is heavily discussed
right now or has been discussed the last
couple of months. Uh the user will in
the morning or whenever this news pops
up, it will pop up straight here and
they can see it straight away. And what
we also built in is a feedback uh since
you know like working with AI agents,
you definitely need that kind of
feedback loop all the time. So you get
them even more adjusted to your personal
um you know liking or what you prefer.
So you can give a quick thumbs up, quick
uh thumbs down, but you can also go give
like additional feedback contest for the
agent. So what do you like about this
news? So what do you not like about it?
And it will learn. So will actually
adjust to each user of the platform.
And if you find uh something interesting
here and you want to provide that
further to the AI agent and have AI
agent suggested content for you, you can
also do that. So either it will act by
itself or you have a human in the loop.
So in this demo I will have a human in
the loop and you can give it an
instruction and actually generate
content from this news. So uh take this news
news
uh article and um
uh change perspective
to our um brand for example and I can
start generating content on that and it
will soon pop up for us.
But then also competitors big thing to
keep track easily on competitors. So you
can also give the agents you know
different kind of competitors or um
different kind of comp companies that
you want to keep track of and say
whatever they provide. So either in
social media or in their on the website
or stuff like that. And same here I can
you know combine different kind of
articles from different kind of
companies and regenerate that into my
own um tone of voice into my own kind of
uh strategy. So I can also you know
either myself as humanly look pick
different kind of channels that I wanted
to provide content for or I can let the
content or the agents um decide that themselves
themselves
>> and is it kind of like daily update or
every hour you know kind of then like
it's updated with the news competitors
and so how often
>> that depends on so some clients actually
have they prefer to have you know news
based on a weekly basis h but I would
say definitely most clients have it on a
live stream. So it updates actually
quite quite a lot uh not every every
minute but every second minutes uh to
get that news straight away straight into
um yeah so that's like the the first
part like media market intelligence um I
would say that you can definitely see
how you interact with AI agents over
time even though it's like it's fine
tuned when you start using it because
it's already learned upon you know the
global knowledge base and everything we
learned with AI agents but uh over time
you also adjust it to your personal preferences.
preferences.
So that's like where we start and I
would say like if you look at bigger uh
marketing communications teams uh this
is like one part that definitely more
like you know planners, researchers uh
sometimes even like H&R um R&D stuff
like that wants access to. Whereas when
we move further and we go into this like
more creative ideas where we have the AI
agents works more like you know some
kind some kind of like um support where
which you can brainstorm with. So here's
more of like a yeah a creative ideas
partner which you can like generate a
lot of content or a lot of creative
ideas upon in a short time and then you
can have the AI agent produce that
material as well. So let's for example h
say that I found a PDF about something
that I think is very interesting and I
want to provide this PDF to the AI
agents and I wanted to from that PDF h
create different kind of uh creative
ideas. So I found a PDF about how we can
scale the global cycling movement and
I'm going to provide that to the AI
agents. I will upload it and then I'm
going to ask it to create different kind
of ideas based upon this PDF.
Depending on company and I would say a
lot of people use PDFs but also videos.
Say for example you have a webinar on 30
minutes or podcast like this and you
want to have like creative ideas on the
just like attach that or a link and it
will uh dig into that context and then
provide valuable ideas based upon it.
Uh so let's say we will want to be a
part of the movement towards
towards
a global
cycling society.
uh base the ideas
on the insights
from the report
such media.
um I start generating ideas uh and it
will in a few minutes you know come up
with quite a bit of ideas and here we
have more you know we always talk about
not quantity more quality but here's
more actually about quantity like bring
up a lot of ideas in a short amount of
time and then the user picks which they
like and which they don't like and I
actually give feedback to the agents
over time so they can also you know
increase the value and the quality in
each of these uh creative
And if we work with a client that
already has like you know they have a
lot of uh good content ideas that
they've used from before we will provide
that to the agent but when we implement
it so they actually also get that value
from start
and once again it's also about giving
the thumbs up and thumbs down here and
giving feedback for the agent to
fine-tune it over time
and um you should also say uh that we
can work with different kind of the
categories here. So some clients have
super strict you know creative ideas
within knowledge sharing or inspiration
or uh conversion uh but here you can
decide that yourself or you let the
agent decide it yourself.
So uh it popped up uh different kind of
um ideas uh and um it has for example
said a howto or tutorial. So, a
step-by-step explainer, carousel, or
video showing how cities like Glasgow,
Santiago planned, funded, and built
their cycling networks include tips on
setting design standards, etc. Uh, and
here, um, you also have a reference, uh,
a way of following references. So, where
did the AI agent actually come up with
this idea from? That was this part of
the report. And this part like when we
talk about a PDF with different kind of
insights that could be value valuable
but definitely when you work with like
more heavy um kind of brief reports on
like your your yearly basis or uh you
know economics and stuff like that which
where the numbers are super uh important
to you know exactly where the agents
provide them from. Uh so the reference
link is actually quite uh heavy usage in
the platform
and then from here some of these idea
the agents can provide and produce
themselves or you actually just take
them on and go back to your marketing
team and you you start like building
outside agent platform. Uh but say for
here we want to generate this content.
It's going to start generating it in a second.
second.
So everything that we produce uh usually
uh pops up here. So when we built the
bike company uh this fictional company
uh we're given some example of bikes
that we have in our sort all different
kind of models and it actually uses
those bikes when it creates these images.
images.
So for example, you can also give it
straight instructions here and it
doesn't have to be made from medium
market intelligence or creative ideas or
say for example uh that I have already
learned AI agents that we have a new
model that we're releasing. I can just
give a simple prompt launching our new
model. I can generate content for that
since it already has that kind of
knowledge about the new model. We
doesn't have we don't have to provide
it. If we hadn't provided two data
agents, I would just add it as an
attachment in the context field
or I can give a simple instruction like
three advices by choosing an electric
white generate content upon that and the
same thing will happen. So if we have a
channel strategy that says that we talk
about these kind of questions in you
know social media on our websites in
newsletters it will provide all those
kind of content pieces. Uh, but it could
also be that we give it a straight
context like here I want to do that on
>> You're missing Reddit which is one of my
favorite ones.
>> That's true. We actually do have it from
some clients and I don't I don't think I
don't think X is on here. Oh, X was on here.
here.
>> But yeah, so but definitely the
important channel. I agree.
uh you know so a simple prompt uh three
advices when choosing an electric bike
uh it will pop up straight here uh and
what kind of like content will be
provided that's al also you know one
common question like will it be super AI
or what kind of quality will it be in
the content and that's where the
reference material comes in so when we
set up a
AI agent for our clients we also use
like reference materials so for this
example when it is a made up company
it's going to be quite you know simple
kind of tone of voice and what kind of
language we use but it can be super
differentiated so we work with you know
B2B companies super big in the recycling
industry but we also work with like
Swed's biggest horse race tracking
company you know BTOC it's all about
horse racing h so it's very individual
and that's like a part of IRA that we
actually work with brand trained AI
agents so it's not within one industry
we work with like all different kind of
and from here uh also like a part that
all our users use is the publish like
the distribution part. So from here we
can also publish all our channels. So we
integrate through you know either APIs
or just partnerships through my or
LinkedIn and such such and we can
actually publish to all channels.
Um it you know when I created IRA I
didn't think it would affect that much
like I thought that the distribution
wouldn't be like one of the actually
valuable most valuable point for the
users but it definitely is it's a part
where people say you know they work with
IR and they would never go back because
they don't want to publish the normal
way like it takes so much time to just
log into each platform you know second
authority methods having you know
WhatsApp messages or uh phone messages
with different kind of six-digit codes
and stuff like that. It makes it so easy
to publish with one one press and it
publishes to all platforms
and also to be able to schedule that
publishing through the platform. So you
can just plan your content for the whole
week and then it will work itself. Um
Um
also one part uh that we usually uh show
up show in our demo is uh how you can
work with the images. So except for that
the AI agents are trained upon like your
brand guidelines and your images. We we
always train with like reference images.
We can also generate content straight in
here. Uh so for example um I can write a
simple to pro prompt generate a photo of
our bike in a Scandinavian setting. I
attach uh one of our bikes and I will
get generated images straight away in
the platform and I can also edit these.
So say for example that I like them but
there's something that I want to change.
Uh, so it's getting colder in Sweden.
Soon it's going to be snow on our
streets. So, uh, update the, uh, weather
to winter landscape.
I can edit it straight away in here. For
example, like brokers working with, you
know, objects that they actually take
photos on like once a year or once a
time and then they haven't sold that the
the apartment or the house. H and then
they're trying to sell it six months
later and it's a different kind of
season season. Uh this is a big thing
for them to be able in like one second
to change know the weather or location
or the e styling of a apartment or
anything like that.
And from here you can also save it
straight into the platform so you can
use it when you publish.
And we haven't launched it yet, but it's
going to be launched very soon. Uh you
will actually be able to also use the
retouch so we can for example just like
uh mark the flowers and take away take
it take them away. You can crop and you
will also be able to provide video very
soon. It's actually in beta. So we have
some users using it right now.
But we can also, you know, just drag and
drop one kind of image or not an image
at all and just like generate a photo uh
of one of our bikes
uh in a um
th European
setting. It should be
one person
and that will also be generated. So we
work with, you know, different kind of
reference images. It's also super good
when it comes to facial recognition. So
you know, put in a person here and um
that kind of person will definitely
appear in the photos and you won't be
able to say that it's not me. like I
have seen like we use it all the time
like you can't tell that it's not me. Um
so it's um it's definitely also a
feature that a lot of people work with a lot.
lot.
>> That's cool.
>> Yeah. So imag generation is um massively
used and it creates a lot of value for
um our clients.
So back to the production and
distribution as I said you can
distribute everything straight away from
here. You can have everything that is
published straight shown in the
platform. You can see uh the links
straight away. So you will have that in
a straight a straight up setting um with
your um image and your um text. You
could also edit of course the text in
here straight away if there's anything
you want to change. For example, uh you
want to add an emoji or whatever.
And then they have we have uh the big
part [laughter]
uh depending on who you are and what
you're interested but you know ranking
in LLMs and also SEO it's been such a
huge uh topic upon everyone's um mind
the last couple of months maybe. Uh so
we train our AI agents both with you
know best practice when it comes to SEO.
Uh if you have kind of like an SEO
strategy, you have keywords that you
rank upon etc. We provide that to the AI
agent but we also optimize them to write
good content to rank in the LLMs.
These uh produ production time for the
articles is about 1 to 10 minutes. I
would usually say eight minutes because
it goes through so many step of actually
optimizing the text. Uh so I provided
some examples in beforehand as you could
see. So for example, write a complete
guide to adventure by bike. uh you get
that article straight away. You can get
all the SEO data in there. You can get
the all text for the images, the meta
description, etc. And if we have that
connection to your website, all the data
will go straight into the website and
you will optimize article in, you know,
we're talking about minutes here. So
it's it's definitely one of those things
that also we have that 95 I'd say 98%
time savings on and a lot of increase in
the the amount of generated content on
articles or in websites than what people
been working in before.
Uh so um yeah it's definitely
[clears throat] one of those you know u
one of those parts of the service that
people use uh the most
and then we have um I would say my
favorite part uh so as I said before
like now I'm a human in the loop I'm
showing off all the steps you can
actually decide if you want the AI
agents to do more of these steps in one
hand automated by yourself so for
example for some clients the AI agent
will pick up media marketing
intelligence uh news and it will provide
you suggested content straight straight
away. So you don't have to have a person
in the loop at that point. You always
have to have a person before
distributing the content. That's the
only like rule we have from IRA because
you know a lot of our branding is in our
clients content. So it's important to
have a person you know reviewing the
content making sure that the text is
okay and the images are okay before
publishing it. except for that the AI
agent can work non-stop from start to finish.
finish.
So uh all our AI agents actually can
provide suggested content. So based upon
the knowledge base and what kind of uh
goals you have target goals uh it can
actually provide what kind of content
you should provide in your digital
channels for this wake and today it's a
lot of social media it's a lot of
articles and not that much like
communications and press releases and
those kind of things but it's definitely
out there uh alive for our clients today.
today.
So an example could be uh through the
planning it knows that we have a new
model uh it's a new product that should
go out there uh the agent has created a
Instagram caption it has created
Instagram images and these images as you
can see super simple but it can be
straight tailored to your graphic design
exactly the way that you want to have it
if you have like in design files like we
will provide that and the images will be
created through that and also I suggest
the publish date. So from here you can
just like give a thumbs up and give okay
and it will uh provide and publish that
content straight away and these kind of
like topics are based upon the client.
So for bike company let's say that they
uh have um they want to both like launch
a product they want to be like you know
share a lot of knowledge and
inspiration. So for here it's a
potential uh Instagram um post about
bike packing in freedom and in emotion
and you have that kind of content
straight away. It's suggest for you. You
can either adjust or you just like
publish straight away from here and it
can be all different kind of uh content.
So you can also you know provide reals
or uh stories or whatever uh format you
want that the platform enables. But it
can also be you know for this example
it's better to just have an Instagram uh
post along with an article and then the
AI agents will suggest that for you. I
would say the suggested content part is
used on the appliance that has it like
80% of the content that the AI agents
provide here people use and then 20%
they might like say that we don't use it
this week and then those 20% instead
they provide uh like instructions or go
through media market intelligence or
creative ideas themselves.
Uh but it's definitely heavily used um
and I love it. I use it a lot for myself.
myself.
And then the la last part also uh I
would say one of my favorite parts but
it's like I'm repeating myself. I love
all the parts about IRA and what the
agents can do but the data and analysis.
So through all these connections that we
have um through the APIs or through you
know the partnership in the
[clears throat] social media platforms
we can also take in the data. So uh I'm
talking both paid and organic here. Uh
so you can actually have a dashboard
that's made up through whatever kind of
targets your or KPI that you have in
your team or individually and it keep
you can keep those on track. So you can
actually you know pull through that data
either on channel level or on the um
integrated level with all different
channels and you can have that delivered
straight away with like live data to see
how you're acting and how you are um
acting against like your KPIs.
And you can also have the guidance to
provide suggestions based on the KPIs.
So say for examples I think some of this
might being switched. I'm sorry about
it. We'll see. Um I'll translate for
you. Uh so for example list generated is
below uh the goal for this year. uh and
the agent will then give you give this
action as as an insight so you can do
action on it and it can also give
insights upon what you should do to
provide a a closer
um you know provide closer like value to
your strategies. to say for example that
you put in a clear goal or topic that
you should increase the social media
followers and you can see that the
increase in those social media followers
are to too low it will give you an
action upon that. So how you can
increase that kind of leverage.
Uh so this would be like this is one of
like the traditional dashboards but I
would definitely say that the data and
analysis is quite um you know user
adapted also because you know every
marketing team they uh keep track on
different kind of things they have
different kind of KPIs
um and it's also based a lot of also the
owners so who's the owner of this kind
of KPI
so that's like the the full platform or
I it's definitely not a full platform
because the AI agents are doing a lot of
different kind of thing other kind of uh
things as well. But I would say this is
like you give it quite a good example of
how you can use AI agents when it comes
to media marketing team even if it's
like a you know small team or if it's a
larger team. Um you can definitely like
adjust how much is working and how much
you're doing yourself and how much the
AI agents are doing.
>> Wow that's this is such a good demo.
Thanks for doing it. Uh okay. So my
question is that
they actually like uh I see that like
you cover all the aspects of marketing
and communication and I see that you
know your eight years of experience in
and has given you that like you start
from the research and you kind of go
with the generation of content and then
distribution and then analysis. So you
kind of take the whole spectrum of this
content and communication part. So my
question is that how did you manage to
build such a platform in in eight nine
months or less than [laughter]
>> how did we so we started off not with
all these parts h so the main part from
start was definitely the production like
producing content and then based like
eight nine months ago it was mainly text
like we didn't even have images from
start and then we just like building
upon that so uh a big uh like um
applause for my partner Lars that's uh
providing a lot of speed in this uh you
know um improvement but I would say like
two parts so one part is my personal
experience and what I've seen that a lot
of clients have put a lot of value in
the last you know eight years h but the
second part is definitely the users so I
have really close contacts still with a
lot of users not on daily basis but
maybe like weekly and monthly and
[snorts] it's definitely a lot of their
insights and a lot of their opinions
that shape what we develop. So for
example this part where you can generate
images outside the provided the
suggested content that's been one of
like the suggestions from our users that
they have asked for h so the last kind
of like four weeks and we developed that
quite quickly have that out in a beta
space to see you know the user action
and how they interact with it and we can
see like oh it's a heavy usage here and
then we need to put more time into
making it better. So yeah, my own
experience but also definitely the users
>> also like I see that um at least the
acco is a bit crowded with it now with
you know like chip PT and other things
like kind of like you should be showing
there um and I see there are lots of
startups coming out so my question is
that you kind of focus in all the like
again you know full cycle research
distribution analysis so do you think
that you know is it this type of
platforms that they do all all the kind
of offerings together wins or is it more
that companies are looking for specific
you know solutions for different
providers like SEOU but then like
production kind of content generation
from someone else so how do you see the difference
difference
>> I think it's an interesting question and
I I guess I don't have the answer
because who has in these times when
everything is like changing so quickly
uh but the way I say like it's more
based on like how users are asking for
and usage of the platform like I think I
personally had more of that part like
you have to have an like you need to
find I need to find our vertical where
will we be like there's so many as you
say just in the SEO and AIO part but
then like when I speak to the customers
they ask for a platform that is full
service like they want to have
everything set up in the same same
platform you know they I just had a
client the other day saying like oh now
I don't have to use Photoshop anymore
because I use all my podcasts front page
in the generated images area and they
still use of course Photoshop for other
kind of use cases but that was like me
like oh you actually find a value here
we can do the full service in one
platform you have to change between all
these different kind of platforms
so I guess that's one part of it but
then I would say like us as a company
definitely don't like we we definitely
see this kind of change uh and we're
definitely open you know with like
partnerships or like finding
partnerships with other companies that
are accelerating their improvement and
their uh you know service within
different fields that we might be able
to provide for our clients within the
platform. So we're still aiming for
having a whole platform and we might be
able to you know pull in different kind
of tools or services that are better
than us because if if we can find anyone
that's better than us and then in in any
spectrum like I am I would love to work
with them like
>> the main goal for me is to have the best
provided content for our customers
whatever way we provide that for them.
>> That's perfect. uh just like moving on
to go to market is that how do you sell
IRA to companies or like what's your go
to market playbook if you can uncover
>> yeah it's um I'm not sure how much I can
cover but to be honest without saying so
too much about what's happening in front
more like reflecting backwards I should
say that we've been actually having
quite a you know slow sales process
we've taken on new kind of clients when
we see that we have the the capability
to do it and actually to have that kind
of personal
you know dialogue with them and we
actually haven't like reached until a
phase where we are now where we know
that the value that provides through the
platform is you know big enough to be
able to scale very quickly. So the path
like in our 12 months or n minus has
been a lot about like face tof face
contact having close relations with our
clients to be able to see the usage to
see the value that we provide and you
know the data on how much and how often
they provide it and it's not like until
now we can actually take that next page
and scale and in what way it's going to
happen I I just like I will have to say
like let's see you you will see in a
short time
>> but I mean like now you know with AI I
see that everyone Someone is very
aggressive on marketing part you know of
course keeping the secrets for
themselves but actually you know like
trying to catch up the numbers very fast
like with 10 million 100 million AR so I
mean as like as a founder who you know
kind of slowed down the process because
of like uh like for a specific reason
like like weren't you kind of you know a
bit feeling you know like tempted that
you might actually you know go faster
because everyone else is faster than you
and then they brag about it and you're
not that fast. So how do you just like
communicate that with yourself and your co-ounder?
co-ounder?
>> Yeah. Uh well I would say like one part
is it's exactly what like you can read
on my page like this is a work of human
result. So for us it's always about
debating and feeling that like you
cannot provide a service for people that
don't stand for this. as soon as the
content doesn't like match up with our
vision that it should be a like a human
result a human quality as soon as you're
out of that I don't think it doesn't
matter like how fast you scale if we
don't have a product that matches what
we uh say that we will match like uh so
that's been like a major part for us to
build upon a level where we know that we
can stand for whatever we provide into
the agents will be good when it comes
out uh so that's definitely has like
holding us back, but it's also been like
should we be just in stealth mode and
you know not showing our brand at all or
should we start being out there a little
bit um and we haven't really been like
ready to take this next step but based
upon that I should also say that we have
a very aggressive uh scalability plan
for the next year and next two years and
I definitely think that this kind of
platform that we are building with these
AI agents it's going to be so much
competition out there so I think it's
going to be like the next kind of like
months upon to like 12 months that's
going to depend on like where we are. Uh
so that pace is definitely definitely
feeling it every single second of my
awakening time but also sleeping times
like you know when you wake up and
you're like da da da like so much is
happening. So uh I'm not scared of it
because I know it's going to happen for
everyone and I'm in the exact same
position. H but I'm very open like open
arms. and welcoming it and we have a
super good go marketing plan uh good
market plan uh which we will address
from 1 of January uh where we you know
launched it in a different kind of way
we can uh also implement a lot of client
a lot faster.
>> Yeah. like kind of like for me like
recently you know I I start like
dreaming a lot about like go to market
strategies you know my kind of like the
solution like what we are building you
know and different ideas and then like
when I wake up in the morning I just
take my notes you know write something
that I should not forget but I was also
asking uh the founders on on X you know
if they're also you know dreaming about
their product uh almost every night you
know about different things maybe
nightmares or like dreams but the kind
of this obsession with the product and
building is like putting the pressure
And in sleep, you know, it comes to your
mind in that like you're working 24/7
actually when if you're sleeping.
>> It's it's uh you know, it's such a
journey. Like I wouldn't do anything
else if I had the possibility. Like I I
love what I'm doing right now. But it's
exactly what you say dreams and then
it's nightmares. It's both parts. It's
247. I have a notebook next to my book
because I wake up quite often like
during nights and I can't remember them
in the morning. So I just like I wake up
and I just type them down and then I go
back to sleep. H but it's both both
parts both nightmares and dreams.
>> Yeah. actually like you know kind of so
I deal a lot with with the founders or
you know kind of founding members who
are like doing the marketing
communication like what I see is that
the perception is here that everyone
wants to keep that keep the secret of
that they use so okay everyone use AI
and you know almost all the content is
AI generated 99% for me you know that
but but somehow everyone is you know
keeping the impression that no it's not
AI it's actually themselves you know
kind of like writing and so My question
is that how do you deal you know with
this customer segment that they're
saying that no they're like no it's me
writing so it's like yeah I not uh so
how do you actually sell this you know
IRA agent and then convince them that
hey guys I know that you're using Chad
PD IRA is much better so please stop you
know fooling yourself uh so kind of how
do you deal with this mindset or with
the sales challenge that you might have
or maybe it's not true or I'm just like
my insight
>> yeah but I can definitely see that as
obstacle in our sales. Uh so it's quite
common especially uh I would space
especially like with people not mainly
working with the content cuz the people
working with the content you know if
we're sitting with content creators or
writers or whatever h they can see the
adjustment and they understand that
there's a difference between like sh mid
journey dal nana banana whatever kind of
um tool that is out there and they can
see the improvement where which you get
in the platform but not only in the
platform also through like references
through the other clients. H so I would
say mainly it's like on a step above
them that [clears throat] people then
work with AI generation in content creation
creation
and I would say like two parts so one
part is about talking about the result
of it. So we've been tracking then
content that has been produced through
AI for 12 months now and we can see and
this is very interesting because a lot
of different kind of you know research
out there we cannot see a drop in the h
result of the creation of content that
we produce and why is that then because
there's a lot of research saying the opposite
opposite
you know I don't I definitely don't have
the answer but one of our guesses and
you know when we've been doing test is
definitely how the um content is
created. So it's so easy today to see
when a text is created through chat chat
GPG for example. Uh because it's just
you know it pulls through all that kind
of patterns. Uh but if you have it
provided through an AI agent you can't
you can't catch those kind of patterns.
Uh so I think that's a main part why we
don't see that drop why we still see an
increase in the results that's created.
And then I would say the other part it's
more about, you know, changing
perspective. So I can see what you're
saying there. Okay. You know, is it's
scary that we have AI producing content
generated content. But if we turn around
and we say what do the users in the
platforms care about? Like do they care
about it? We can still see that the you
know interactions in a lot of posts out
there have like increased so much. So
yeah, it's a lot of um I was going to
say bad word, a really scrappy content
out there. I agree. I don't love it. Uh
but we can also see a lot of like high
quality content that's produced through
AI. So I'm always always trying to like
know have them change that perspective
and see the other way around.
>> Yeah. But also I feel that it's more
about quality and also quantity, you
know, like in the content like
production. So you really kind of need
to push a lot because you don't know
which content would go viral. So you
just have to try out everything and then
as a human it's impossible to think of
you know novel ideas every day. So it's
better to optimize AI and kind of like
like get a nice overview of the research
and everything else and then put
something more novel because as a human
how much can you consume information you
know so it's impossible. Um,
>> I agree with you so much there. Like
quantity is a part of this and I it's a
lot of people that don't agree with us.
Uh, but I think that's a major part of
it. Like you need to have that quantity
out there as well because that's the
landscape today. Like you have an
increase in how much content is produced
out there and you need to keep track on
that. Like you can't stay with like oh
we have one uh social media post a week
or we have one article a month. Like
that does not work these days. Yeah, one
question about the SEO part in AI, you
know, is SEO different than artificial
intelligence? Um, like you call it >> optimization.
>> optimization.
>> Yeah. Optimization, AI, whatever we call it.
it.
>> Yeah. Okay. So, I should have Lars here
because he's our SEO guy. [laughter]
He's both working with like, you know,
uh, front end and back end when it comes
to SEO and he's the guy when it comes to
AIO as well. Uh, so I'm always just
trying to, you know, use his words when
I explain this part. Uh but I would say
like yeah you can see difference like
there is differences in how we optimize
content through um for Google or Bing or
LLMs exactly how those are done that's
also like always one of those questions.
So then how should should we do it? Uh
and that's a part of the journey to you
know adjust and test and also have good
traction in how you calculate and how
you follow how you are mentioned in LMS
and how you rank. uh and that's like one
of the part where we have partnered and
we've partnered up with another company
to you know keep track on how our
clients are doing in this kind of
ranking. Uh so it's also for us it's
also a lot of like keeping track of what
actually is happening out there and what
is working.
>> Perfect. Perfect. Great. So reaching out
to our last part I'm going to ask random
questions and I would love to have your
answer on that. You have been building
agents. Yeah like with AI. Uh so now
what agent or what AI has taught you
about humans?
>> What has AI or agents taught me about humans?
humans?
Uh such a good question. Uh like one I
don't know if this exactly answer on
your question but I would say like one
part is definitely like the human
behavior. Uh so working with AI and AI
agents I couldn't understand how quickly
I say it in two perspectives how quickly
people adjust and then get used to using
AI agents. Like I had I wrote a LinkedIn
post about it yesterday. I had a client
that actually h she has she gets five
generated AI photos for each post and
she asked me to decrease that. And I was
like why do you want me to decrease it?
And she was like, "Because I have to
delete three of them." And she was like,
"That takes a few seconds." And I was
like, "You care about seconds." Like, I
love that, but that's definitely a like
change in the human behavioral. H. So, I
would say that's one part. And the other
part that AI and AI agents have taught
me is
um you know it's the easiest way to
change behavior and like to get people
to get that usage up in a high levels
quickly is to have them use it
themselves. Like you can talk and you
can talk and you can show presentations
or do demos but it's like they need to
be in there and they need a lot of them
need to have that still safety like the
human safety to feel that they have
someone next to them because it's still
scary like can I do anything wrong and
even if I'm saying like you can't
publish anything through your shadows
because we don't have any kind of
connections etc et
people are still like insecure um which
I think is also part of the human
behavior and I guess I was the same a
few years ago when I had like when I
weren't used to this technology.
So does that answer your question? Yeah,
of course it does. And like I mean
reflecting on that yeah kind of also for
me humans you know surprise me a lot is
that you know they are shocked oh wow
kind of you know AI came and now
everything is new kind of like humans
are not I mean like they're like like
useless but after a while you see that
okay they're actually you know expecting
more from AI and then like they find you
know use cases where AI cannot do it. So
I think that humans always you know kind
of expect more and so like you can like
like never say that actually actually
yeah like this is going to be it like
with AI so like humans evolve faster and
um no that's kind of yeah like uh I ask
a lot of like founders who are building
AI agent because I think that building
an AI agent or a tool it's kind of
taught you a lot about how humans think
and how humans behave and it shows you
like so much more because you kind of
want to build the hum like like digital
version of a human with that agent
>> and you kind of select like trend that
and then see it. So my next question is
that you know building a startup um
brings you nightmares, nice dreams, you
know lots of working hours, execution,
stress. So uh how do you normalize your
life as a founder you know kind of how
do you keep your social life there the
families health uh good sleep combined to
to
>> do you that's my question back to you do
you normalize
>> they normalize it at all [laughter]
>> yeah I guess
two parts I would say like one part of
it is actually having your social life
and your surrounding to be aware of your
journey Like um so I had my husband and
my 2 and a half yearear-old daughter at
home like we definitely had like a lot
of discussion on me going into this cuz
I was like well this is going to change
a lot for you like it's going to be you
that picks her up, drops her off, you
know, put her to bed and I still spend
time with her of course but it's like it
it definitely affects a lot of your
personal life and if you have your
surrounding on board on that it's going
to help you so much. It's going to give
you much less stress. So I say that's
one like big part and then I would say
the other part like so all around me I
always had entrepreneurs so my brother
has his own company my parents had their
own company for me it's like um it
doesn't have to be wrong to you know
live 24/7 with the company like I'm
happy doing it like I love my job and I
love my life so you know why should I
change it and then of course like care
about your hours of sleep and care about
getting training in there and make sure
that you fuel your body with like good
food and nutrition and stuff like that.
Like I care about that those things a
lot. Um and I prioritize them in my
daily basis very much. Like ask anyone
in nursing and they would agree. Um but
then like also just accept that it's a
journey that you're on and like if you
love it then like accept that it's that way.
way.
>> Yeah. like you accept uh all the bad and
the good ones of it and like it has both
of them and just like takes more time to
like reach to the good ones uh at some
point. But um so my next question is
that what do you do uh on the weekends?
>> What do I do on the weekends? Uh I spend
a lot of time outside.
>> What do you do the weekends?
>> Not working. [laughter]
>> Uh it's definitely work during the
weekends but not all hours. Uh so I
would say not sleeping. I don't sleep in
like I love going up at 5 and then I
have my hours of work before my family
wake up. It's the perfect morning for
me. H but then usually you know I spend
a lot of time outside that's recovery
for me like being out in the forest you
know having lunch in the Swedish
settings beautiful nature here. Humm
and then you know training is also
essential part of me. I do crossfit so
definitely have one to two CrossFit
classes each weekend h and spend time
you know with friends and family. So
trying to put that into and then having
those hours in there which don't you
know um doesn't affect too much of the
family life.
>> Yeah that's this going to be my last
question because I just like remembered
it is that as you're a first- time
founder what are some of one or two
advices you know for first- time
founders who are just starting now to be
mindful of. Yeah, first part would be uh
just an advice I got from um to Ben at
Antler. Um so he very quickly gave me
advice like don't rush on that partner.
Yeah. like you need him cuz your whole
company is built upon a technical
perspective and knowledge. But like
don't rush on finding that person
because if you find the wrong person
that's going to give you so much
headache and I'm very happy that I gave
that time and uh I was okay with that.
Um, so I would definitely send that
further away like or to the next person
like don't stress like you need to find
the right person and you can't do the
journey yourself but uh it's better to
have the right match than just picking
someone in a fast pace. I would say that
it's my first advice and the definitely
the second one is be prepared like be
prepared prepared in the sense that you
work with this day and night like you
have nightmares and you have dreams and
it's a lovely journey and if you're
thinking of doing it like do it like
it's amazing but just be prepared and
you know give the surrounding and your
family or friends like a heads up like
the next couple of years is going to be
different if you haven't done it before
of course. Yeah, it's so nice where to
end this episode. Just be prepared and
like kind of like like be ready.
>> So easy.
>> Yeah, it was super fun having you like
meet me over and thanks a lot for for
the chat and I hope to see you in person
as well soon.
>> Yeah, same. Thank you so much for having
me. I I really like this hour with you.
>> Thank you. Bye. Bye. >> Bye.
Click on any text or timestamp to jump to that moment in the video
Share:
Most transcripts ready in under 5 seconds
One-Click Copy125+ LanguagesSearch ContentJump to Timestamps
Paste YouTube URL
Enter any YouTube video link to get the full transcript
Transcript Extraction Form
Most transcripts ready in under 5 seconds
Get Our Chrome Extension
Get transcripts instantly without leaving YouTube. Install our Chrome extension for one-click access to any video's transcript directly on the watch page.
