Ep. 552 w/ Burhan Fatah Chairman & CEO at Sivoo
Kevin Horek: Welcome back to the show.
Today we have Burhan Fatah.
He's the chairman and CEO at Sibu Burhan.
Welcome to the show.
Burhan Fatah: Glad to be here.
Kevin Horek: Yeah, I'm excited
to have you on the show today.
I think what you guys are doing at Sibu
is actually really innovative and cool.
You also have a really
impressive background and you've.
Done some pretty incredible
things, especially with your
early days in the internet.
Um, but maybe before we get into that,
let's get to know you a little bit better
and start off with where you grew up.
Burhan Fatah: Sure.
Uh, so, uh, I finished,
uh, my education in Europe.
Uh, okay.
I was born in Pakistan and then I
came to Europe for higher education
and I finished my studies in physics.
Actually, I got a PhD in physics.
In particle physics.
And those early years in my career, they
were instrumental because I got to meet,
uh, some incredible people and I also
got to work at some incredible places.
One of those places being C and
oh, interesting son is the largest
particle physics lab in the world.
And during my student years, I
got to meet Tim Burners Lee there.
And, uh, he created an app for scientists
to communicate among themselves.
I was fortunate enough to use
that application in certainly age
early days, and that application
eventually became the worldwide web.
So, uh, my involvement with Worldwide
Web goes right from the inception.
Uh, Then I moved to the US and I
switched to more computer industry.
Uh, and in that, in that capacity for the
past 20 plus years, I've been involved
with the various types of companies,
uh, online media and attainment,
social media, cloud, mobile computing.
And, uh, it management.
Uh, so, uh, you know, within
that capacity, uh, part of my
job was, uh, work, uh, was on
Sibu, uh, the parent company of
Sibu, uh, which was broad relay.
I started that company back in 2000.
And, uh, that, if you
recall that part, uh, time.
Uh, at that time, uh, Broadband
was just introduced, uh, sure.
By, I believe, Verizon before that
19 up to 1999, we were still on the
Dialup and, uh, video obviously won't
work very well on Dialup connections.
And, uh, so, uh, that was the early days.
Uh, we practically have to
invent, uh, develop everything.
There was no open source software,
there was no compression technologies.
Uh, there were not even video
standards, very well defined.
Uh, So for the first few years between
2000 and 2005, we mostly, uh, worked
on developing new technologies.
Uh, the initial focus of, uh, uh, my
company was to bring international
programming, uh, uh, To the US audience,
US was the only country which was
connected, uh, over high speed internet.
So we brought international programming
and, uh, distributed streamed in the us.
So that's, uh, my background, uh, in
terms of, uh, uh, entertainment, uh,
and media, uh, I mean mostly online.
Oh,
Kevin Horek: fascinating.
Okay.
So walk us through.
What Sibu is today, because there's been
a lot that's happened on the internet,
nevermind just the video space in,
you know, that period of time, right?
Yes.
Burhan Fatah: So, yeah.
Uh, I mean, so the internet, I
have seen pretty much the entire
internet, uh, evolution, uh, starting
2000, which was a banner year.
Yeah.
And until now, and it has gone
through a lot of iterations.
Uh, as I said, there were not a
lot of technologies available,
uh, at that time, so we had to
pretty much, uh, invent everything.
Um, content was not readily available
because, uh, content providers, they
were not sure if, what is this new media?
Uh, uh, it was very hard to get.
Yeah.
Exactly.
Sorry.
Exactly.
Yes.
Uh, uh, I'm sorry.
Uh, so yeah, it was very hard to get
content, but, uh, you know, after a lot
of convincing and talking to various
content providers, they used to attend
a lot of conferences to acquire content.
Uh, in initially they were very
reluctant to give us any licensing
rights, but after they saw few times,
then they got more comfortable.
By the mid two thousands, the
word was out as well, that.
Uh, video streaming is, uh, you
know, becoming more ubiquitous.
And so that allows, allowed us to
start licensing professional content.
Uh, and, uh, we ended up having
actually one of the largest libraries
of online content distribution by 2005.
Wow.
Uh, at that time.
And worked with some incredible
companies, uh, at that time.
Uh, and, uh, Um, you know, especially
work with the, with, so in 2006, Amazon
introduced, uh, uh, the public cloud.
I mean, that was the
first public offering.
So we were in that cloud
like pretty much right away.
Uh, that was my first
introduction to a public cloud.
And uh, uh, you know, we
were like the first video
streaming company in the cloud.
And, uh, so that was,
uh, a great relationship.
Uh, it's hard to believe where AWS
is today, uh, compared to 2006.
Um, they had, uh, absolutely no
service, no instances other than,
uh, just their elastic computing.
E C two.
They didn't even have s3, which
is the database, just one app.
Interesting.
And today they have over 200 apps.
And, uh, you know, and it was a very
initial stages of that, uh, that program.
Uh, then what happened in 2007, uh, when
Apple introduced, uh, a iPhone, we were
looking into mobile even before that.
So that was a major, major revolution,
uh, both in terms of internet connectivity
and in terms of video streaming.
Um, Netflix, uh, then launched it.
Okay.
Before Netflix launched, uh, there
was, uh, obviously a recession
in 2000 end of seven eight, which
affected a lot of companies.
There was a lot of activity in
video streaming at that time.
Right.
And that affected the
entire industry a whole.
Uh, series and bunch of
companies even very well funded.
They went out of business.
Uh, that was a very tough
time because nobody was really
generating any revenue at that time.
It was just, uh, some kind of market
creation phase adoption, user adoption
phase, and that turned out to be.
The biggest blessing, uh, for
example for Netflix because they
launched right around that time.
They switched over, didn't
exactly switch over.
They also added video streaming
alongside their DVD business.
And, uh, then in 2009 on the other side
of the recession, the whole industry
started getting, uh, momentum, but there
were hardly any players left because
most of those players were wiped out.
Uh, we also had a lot of setbacks,
uh, during that period of time.
Uh, and, uh, so the, the market
then from there on it started
developing quite a bit more.
And I would say between 2009, 10
until today, this is pretty much
the golden era of video streaming.
Uh, We have seen this progression.
Uh, uh, other countries caught up.
Uh, more services became
available in many other countries.
Uh, US started picking up more, more
and more companies getting involved.
Hbos and Hulu came into
thousand 11 and so on.
So this thing, uh, pretty much, uh, you
know, picked up, uh, since that point.
So, ne fast forward to today and,
uh, Today is a very interesting
market and, uh, we had to change
our business model accordingly.
Um, video streaming was pretty much
concentrated until few years ago in
the us and as a result, all the global
bigger companies, they were all based
in the us whether it's, uh, you know,
I just mentioned few of them, whether
it's Netflix, whether it's Disney.
Um, Hulu and, uh, you know, paramount
hbos, the list goes on and on.
Even Amazon Prime we can
include in that category.
Um, but what happened is US market,
uh, became extremely saturated.
It is very saturated market today, more
than 80% of the households, uh, today
in the US they have at least one video
on, you know, video streaming, uh,
subscription, many households, multiple.
Um, so the opportunities are getting
harder and tougher to add any more users
because pretty much anyone who needs a
video streaming service, they have it.
The situation is very different
actually on the international
scene, uh, in many other countries,
especially emerging markets.
Uh, those markets survey
US was 10 years ago.
And they are pretty much at
the early stages of adoption.
Uh, it's, it's, it's, uh, growing much
faster than the US market grew because
there were a lot of other issues, uh,
infrastructures and, uh, you know,
smartphone, uh, penetration and so on.
But today, if a smartphone is in the, in
the hands of over 5 billion people, uh,
across the globe, that's where they get
most of their broadband, uh, services.
And as a result, uh, the video
streaming is except growing
extremely fast in many countries.
And we are, uh, lucky
to be in those markets.
We switched over our business model.
We are still have a, a good presence in
the us but bulk of our focus right now
is in emerging markets in Asia, Latin
America, uh, and Africa, of course.
And, and we, we are, we are
seeing a tremendous growth
in many of those markets.
Um, our.
We have millions of users, uh, currently
in, uh, roughly about a hundred countries.
Some countries is very small footprint.
Other countries.
We have a lot bigger presence and the
way we distribute content today, uh,
we are, uh, We are essentially not
doing retail or direct to consumer.
We work with global telcos, uh,
who are the gatekeepers of the
internet, uh, mobile operators,
uh, ISPs, if there are any left.
Uh, and they become our
distribution platforms.
We have, uh, agreements with them.
We bring content, we bring technology, and
they give us access to their user base.
And that seems to be
working exceptionally well.
Kevin Horek: Okay.
So let's dive a little
bit deeper into that.
So like, uh, like Verizon, for example,
would build their own video streaming
app and you would provide them content.
Is that correct?
Or
Burhan Fatah: how does that work?
Well, we are not working with Verizon
today, but we did work with Verizon.
Actually, Sibu was the first service
when they launched their I P T V service.
Uh, uh, Verizon Fires when it
was launched back in, uh, 2006 or
seven, I believe that timeframe.
Right.
We had the first three international
channels and Verizon fires.
Oh, cool.
Yes.
And uh, but then for, uh, because
of recession and all that, Verizon
fires didn't really grow very well.
Okay.
And speed was very slow.
Today we are indirectly distributing
content through Verizon, and I'll
get to that, but we are not providing
any content today to Verizon.
But we did provide, uh, hundreds of
international movies and television
programming to Verizon's in the past.
Today they have their own I P T V network.
Uh, they have their own users over fiber.
Since our focus is not.
Not too much on the US markets right now.
Right.
Uh, but there is absolutely, uh, we
are going to be working with 'em.
They are distributing some of our
content today, uh, on, uh, but on a
different platform, not, uh, on their
own, uh, a platform, which is I P T V.
It's a different platform, which
I'll talk about a little later.
Kevin Horek: Okay.
Very cool.
So how, well, you've obviously
been doing this a long time.
You've had to pivot.
Probably a number of times, and
maybe not, maybe not just like
full pivots, but small pivots.
And you've been through,
you know, some recessions.
How have you stayed relevant and adapted?
Because I think a lot of companies
are going through that right now.
Burhan Fatah: Yes, exactly.
Uh, right now, this is actually
a pretty tough market, uh,
for online video in general.
I mean, we see.
Almost all companies are affected by
that, and there are reasons for that.
And, uh, if you want me to go a
little deeper into those reasons,
I'll be more than happy to.
Sure.
Um, what's happening is that most of,
as, as I mentioned, most of the bigger
companies in the world, they were.
Mostly based in the US and they
didn't really go out that much.
I mean, uh, even companies, uh, such
as Hulu's, they have strictly just,
uh, USA based and Paramount and, you
know, uh, most of these companies,
Netflix, uh, has gone international.
Um, but bulk of their user
base is still in the us.
Disney tried a little bit international.
Uh, they are kind of, I mean, they
have other issues they're dealing
with, so this has been a very tough
market because everyone is trying to
compete in a market which is saturated.
Right.
So that's one of the reasons why these,
uh, companies are having issues increasing
their revenue base or user base.
Number two is that the expectations from
some of the larger companies was, uh, that
they were going to make a lot of money
from video streaming fairly quickly, and
that it didn't turn out to be that way.
Um, video streaming obviously is cheaper
than a traditional media, uh, where it.
You got your feet from cable channels and
so on, which were very lucrative business.
Uh, video streaming revenues are not close
to those numbers even today in the us.
Um, I think.
International markets represent
the biggest opportunity.
Think about this way, U s A has about,
uh, 3% of the global population.
97% of the people live outside the us.
We have about 300 million phones out of
over 5 billion smartphones in the world.
So that means majority of the people
who are potential customers, candidates,
or who can use video streaming
services, they are international.
Now there are major challenges of going
international, especially for many
American companies who are very used to
charging high payment subscription fees,
as well as dealing with the, I mean, uh,
you know, uh, cultural issues and so on.
They're very familiar with.
Going international poses a whole
new challenge and I've been studying
international markets for being an
international, having myself international
background, uh, it's easier for me.
I have lived on four continents, so I
understand the cultures speak several
languages, so I can relate to many
of those things and I can see where
the problems are and we are trying
to solve some of those problems.
Problem number one is obviously the
cultural issues and the language issues.
When you go to Africa, Asia, almost every
country speaks a different language,
so you have to be pretty comfortable.
There is still a lot of government
bureaucracy, red tapes and so on, so
you have to get through that process.
Number three, infrastructures
are still pretty bad.
Uh, they're not as advanced, developed as
in the US and Western Europe and so on.
Many emerging markets, they're, they're
working very hard and developing very
fast, but they're still not as developed.
And number four, which is the
most important point for many
of these providers, is that the
price points are much, much lower.
Um, in us, uh, you can offer
a video streaming service.
Netflix, for example, they can charge $15.
Most other countries emerging
markets, $4 will be considered
as a high subscription fee.
The, the issue then becomes, Because
the way many of these bigger companies
are set up, it's very hard for them
to offer services, high quality
content, offer two $3 a month
subscription and still be profitable.
And that's turning out to
be the biggest challenge for
their, uh, penetration globally.
We knew these things, uh, several
years ago when we started working
on this project going international.
And the way we constructed our business
model was that even if we sell a
subscription, say a couple of dollars per
user, or if we are making money throughout
supported content for a couple of
dollars a month, we are still profitable.
Business model was built that way.
Uh, another challenge was the,
the infrastructure challenge.
And in order to overcome that
infrastructure challenge, we made,
uh, a decision early on about three,
four years ago that we are gonna build
our own CDN for global distribution,
essentially building our over our own
private cloud to distribute this content
to the far reaches of the planet.
And that strategy turned out to be a
great strategy because, That allowed us to
partner with global telcos, bigger telcos,
many telcos have lot more users than
Verizon's or at and t in the us right?
For example, China had the two
largest telcos in the world.
Um, India has the third
largest, uh, telco, which is
Geo MTN in Africa have 200.
They have 250 80 million users and so on.
That's a large user base.
But the problem is that a lot
of these telcos are local.
Many of those telcos are not global, so
they don't have any presence outside.
Their countries are outside their regions.
So the question then becomes how do you
deliver this content to these companies?
And yes, you can use, uh, you
know, traditional, uh, method herd.
You can, uh, transfer this
data over a public cloud.
You can provide satellite feed and
all that, but that is cumbersome.
They can get very expensive and you don't
have any control, especially some of the
emerging markets where you're going to.
So our cloud idea ended up become, uh,
the big winner because we were able
to deploy our data centers very close
to where these telcos are stationed.
Or where, where they are based at.
And we were able to directly
deliver, connect our data
centers to their data network.
And that way there was no latency.
As you know, latency is the biggest
killer in video distribution.
Few milliseconds can make a huge
difference, uh, because that degrades
the quality of the video and nobody
wants to see the buffering and all
that, you know, low quality video.
So that strategy is working very well.
Uh, we today have four data centers
on three continents, and our goal is
to expand those data centers into 50
countries where we work, uh, where we
will work with our telco partners and
so on to distribute, uh, this content.
Kevin Horek: Interesting.
Okay.
So how do you actually get the content,
and then how long does it take to
actually reach all those data centers?
Is it like.
Minutes, hours obviously probably depends
on the size of, you know, if it's a short
form clip, if it's a TV show, or it's
a movie or, or walk us through that.
I guess that's kind of
a two-part question.
Burhan Fatah: Okay.
No, it's a very good question.
It goes to the heart of the matter.
So yes, con, without content, there
won't be any video streaming, right?
So once we get that out of the
way, so what happens to content?
Yes.
I mean, you can create your own content.
Or you can, uh, License content.
Uh, there are a lot of
outlets which produce content.
Hollywood is the largest content
production factory, I mean, on the
planet, uh, and other countries.
Similarly, Bollywood and India.
And, uh, you know, Japanese,
they have a pretty strong, uh,
content, uh, markets and so on.
So almost every country has,
uh, some content, at least
their television programming or
movies or whatever they produce.
So, We are not into content production
business, at least not as of this moment.
So we license content.
Uh uh, I've been personally
licensing content for over 20
years, so have a lot of experience
in negotiating deals for content.
Sorry, can I ask you if they can get
pretty complicated, especially if you
Kevin Horek: Yes.
Sorry, I was just gonna
interrupt you there.
So how do you decide which content though?
And then do you send the same content
all over the world or how does that work?
Burhan Fatah: Uh, no, actually the answer
to that is, uh, a very complicated answer.
Okay.
And con yes, you cannot take
the same content in the same
language and distribute across
the globe because that won't work.
Uh, content has to be localized, uh,
into, at least into the major markets.
Right?
Okay.
Uh, just to give you an example, uh,
Yeah, let's say we take, uh, I mean
our content, uh, we licensed from
Hollywood and we take it to India.
Okay, so 16% of population
of India speaks, uh, English.
So.
And, uh, but other 84 doesn't,
so they won't be able to
understand that content and so on.
So just, I mean, you know, just,
uh, streaming, uh, uh, English
content in India won't work.
I mean, you may have a small user
base, but not, so that means that
is essentially you have to, so there
are two possibilities in this case.
One is that, uh, you acquire
content for all your major markets,
specifically for that market.
Uh, India, for example, you
acquire content in Hindi.
If, uh, you're going into Middle East,
you acquire your content in Arabic.
If you're in Africa, you
acquire content in Swahili.
And, uh, you know, there
are other colonial times.
Many of the languages are also spoken
English, French, uh, and uh, some
Spanish and Portuguese as well.
So, So that's challenge one.
You have to acquire lots of different
types, uh, I mean language based content.
That's challenge number one.
Challenge number two is that you
have to mesh that content with the
viewing habits of that country.
Some people, uh, countries are more
interested in tele novellas, soap,
soap type of TV series, other like
movies, other like sports and so on.
This is all content.
So that's a second match.
Uh, we've been working on
solutions, uh, Kevin, for many of
these, uh, situations for years.
Uh, we have a pretty
large, uh, content library.
I mean, uh, Um, over a million videos
of files and, uh, many of, we have
over in 16 languages and we are
constantly fine tuning, uh, how to
match content for different markets.
Got it.
So makes sense.
That's, uh, that's the content.
Yes.
Kevin Horek: Okay.
And then I guess, so once you decide on
which content, so then for your, the CDN
and your content delivery network, do you.
You only send the proper content
that's relevant to that country,
that DA data center, or do you send
everything to everything everywhere.
Burhan Fatah: Okay.
So, uh, that's again a great
question and again, it doesn't
have a very simple answer.
Sure, fair enough.
Um, Uh, content, uh, funest content,
and we delivered an entire system,
which, uh, this is our proprietary
technology, which deals with the
entire life cycle of content.
What does that mean?
So you get content into your system.
Let's say we license content, uh, from a
player, a let's say somebody in Ireland.
We license content, okay?
Now you need to bring this
content into your system.
So number one challenge is
the content ingestion part.
How do you bring that
content into your system?
In the old days, when I originally
started, they used to be VHS tapes, right?
And uh, so people just send you a
whole bunch of a box full of VHS
tapes, and from there you get the.
Uh, then it was D V D and it
was a little easier to ship more
content than the hard drives.
If it is content, you get on a hard drive.
Then when the internet speed grew a
little bigger, you can do file transfers.
Then they were satellite feed
you can get from transponders.
The idea is to get
content into your system.
Right.
So let's say we got content into
our system, which we get, we use
many different ways it comes to,
into our system, but that's just the
beginning because that content has
to be distributed across different
platforms in different places where
the network speeds may not be great.
And, uh, you know, many other,
I mean, uh, considerations.
So first of all, when we get content,
uh, we can word that content,
uh, depending on the country.
Uh, First into a flat file.
And that content, first of all, has to
be encoded in what pla what avenue we
are gonna be using for distribution.
Is it a mobile platform?
Is it a web platform?
Is it an O T T service?
So content has to be, uh, adopted
according to that platform.
And then you next you see, is the
network connection, is the speed, is it
their great connection or is like, Not
great connection that will determine
whether they're gonna watch a standard
def, uh, video, or they're gonna watch
a high DEF or 4K or eight K because
the file starts getting bigger as you
start moving to higher resolutions.
So those are the next decisions.
So once those things are done, this, all
this process happens within our network.
Once we got the content, so we
converted accordingly, weathers.
PAC four stream, whether it's H 2 64,
H 2 65 i TP tv, whether it's cable
apps, specs, all that things take
into consideration so that that is all
automated process within our network,
which happens once that is done.
Now the next question is how
that content will be distributed.
So we need to now get this content
in the hands of the end user.
So the way that works is, and that's where
our data center strategy comes into play.
Um, we have already strategically placed
data centers in Asia, Europe, and u s
a and we'll be adding, we are adding
more and more nodes, uh, amend a lot
of nodes planned even for this year.
So then we bring that content to the, and
by the way, all these nodes are connected,
so this creates a private cloud.
And the speed of transmission with
the cloud is, uh, milliseconds.
Okay.
I mean, it's very, very quick delivery.
Even content ingestion, just to give
you an idea, one of our providers
is in Ireland, so the content
leaves Ireland and uh, it goes to.
We are again, changing our, one of
our ingestion facilities in Vietnam.
That ingestion is done and then
that content is transferred over to
the us uh, uh, to our data center.
And once we have done the rest of
the processing, which I mentioned,
then we forward this content.
To our data center in Frankfurt.
From there, it gets delivered, for
example, to Africa because we have a
direct connection with our partners,
African partners, right from Frankfurt.
We connect with this now, if you see
it's about three or four continent
movement, which is almost happening in
real time, I would say, from Ireland.
With the content originated to
delivery in Nigeria, going through
these three, four continents in the
meantime will take few milliseconds.
Wow.
Kevin Horek: That's
in, that's fascinating.
That's cool that you can do it that fast.
And then that includes, and just, I think
you mentioned this is, so you, if I send
you like a 4K video file, you'll cut it
down to 10 80 p, optimize for mobile,
optimize for my television, and all the
other different kind of form factors.
That's
Burhan Fatah: correct.
That's correct.
Okay.
Uh, right now we are running, yes.
I'm sorry.
Go ahead.
Finish your thought question.
No, no, keep going.
Yeah.
So right now we are running seven apps.
Okay.
Seven different apps of distribution.
Uh, the, the, the apps
you're familiar with.
I mean, Android and iOS and web app.
I mean, they're accessible in the
US and anyone can access them.
Um, the other four apps are running for
platform, which are not known in the us.
Three of those platforms are in
India, and they're mostly running
on, uh, different types of phones.
Those phones are, they
are a semi smart phones.
They are built for the masses.
They're inexpensive phones and,
uh, But you have, this is a total
different format and uh, so you
have to convert into that format.
And then similarly, all the African
distribution, we are working with
the, the world's third largest mobile
operator who also run an operating
platform, which is totally different.
Uh, and so we have to encode according
to that platform to distribute it.
So yes, I mean all these, uh, seven
platforms, uh, being distributed, uh, and
they cover most of the online streaming,
uh, you know, part of the delivery.
Got it.
Kevin Horek: Okay.
Very cool.
So how does AI play into all this?
Or does it at this point?
Burhan Fatah: Okay, that's
another great question.
Uh, I've been looking into
AI for many, many years.
Sure.
Um, a, a very fascinated by ai
and, uh, AI is not a new concept.
Ai.
First time it was, uh, mentioned
was, uh, in the fifties, uh, by a
Berkeley professor, uh, who kind of
defined, uh, a very base level ai.
So AI has been going through
tremendous r and d in the sixties,
seventies, eighties, nineties.
And uh, the question then
is that why all of a sudden.
AI became such a big deal
in a matter of months.
Part of that credit you have to
give to Chad g b t on November
30th when they introduced to every
person a large language model.
It, uh, every person realize that
regardless of their knowledge,
even if they have zero knowledge
about computers, they can still
go and do some fascinating stuff.
So that created a massive awareness,
but it didn't happen because of that.
Whenever there's a new technology,
um, when it's talked about, it takes
many, many years, in some cases,
decades before that technology
becomes, uh, ubiquitous or it's adopted
by the, you know, by the masses.
Yeah.
The reason for that is that a lot of
those things are futuristic when a
technology is defined and all those
things have to come into place at the
same time for that thing to happen.
This thing exactly happened in
AI space in the last year or two.
Uh, Chad, g b t was the straw, which broke
the camel's back, but all the perfect.
Conditions were there.
And what, there were two main conditions
for AI to be, uh, useful or functional.
Number one is your processing power.
You have to have an incredible
processing power because the concept
of AI is, the big main concept
behind AI is called machine learning.
That means your computer or
device has to learn all the
variations within that framework.
Um, so.
That is a learning process.
Think about it as a baby, a newborn baby.
A baby doesn't know anything, baby.
The moment baby opens his
eyes is start learning.
It looks around it, see, this is light.
This is eventually sooner start
realizing, this is my parent, this
is my mother, this is my father.
And over time, that baby has
become a teenager and grownup,
it learns all those things.
That's machine learning.
So now the now, An adult knows many
things because they learned over this
timeframe through trial error, error,
error, and through experimentation and
just learning from the environment.
AI functions in very similar fashion.
So you need to train your AI product
with all the possibilities in all the
vari variations so your computer can
give you answers which you understand,
or which you want it to accomplish.
So that's where ai, so now ai, uh,
the horse is out of the barn, and
now I think there is no looking back.
And, uh, the AI, in my opinion,
is the biggest revolution
since the internet itself.
Um, interesting.
Uh, yes, mobile was great in 2007, but
AI is gonna for, uh, supersede that that
industry, uh, the reason for that is,
That AI is not something you don't have
to do at the user level or as at the
business level, anything very different.
This just sits on top of
your existing systems, right?
And it creates efficiencies
across the board.
I mean, so if you can create more
efficiencies without changing your
infrastructure, that's a great deal.
This is literally like an aby sits
on top of everything else you have.
And, uh, and, you know, and
it creates efficiencies, uh,
which goes to the bottom line.
So you're profitable.
Uh, I can think of a single industry
today, which is not going to be
a user of AI in the coming years.
So after that, obviously
a big proponent of ai.
Now we have to see how AI fits into our
business model and what we can do with it.
So ai, we can do a lot of, uh,
fascinating things with ai.
Uh, we can do, for example, AI
dubbing, uh, I was just mentioning
a big challenge in distribution of
content, uh, across the globe because
it has to be in the local languages.
How about we get a large library
from Hollywood and dub it
into 10 different languages?
Right.
Um, traditionally dubbing was a very
expensive, and then we can dub the English
content into French, German, Spanish,
Portuguese, Hindi, Chinese, Swahili.
That'll cover 80% of
the global population.
But the traditional problem with dubbing
is video dubbing, is that, uh, its,
uh, Very expensive and time consuming.
You have to get all these orders to
actors into your studios, record,
re-record, rehearsal, all that
stuff, which goes into it, and it
becomes very expensive, actually.
Hundreds of dollars per minute.
This is the cost of dubbing, so
that makes it almost impractical for
dubbing large quantities of content.
Here comes ai.
You can train ai, you can
create a model, you can train
ai, so it dubs automatically.
Content you just feed it into, okay, this
movie and it spits out in other languages.
Uh, very initial stages of this
technology and its application and
dubbing are already, you can see it
today, I believe within a couple of
years you are gonna see, or maybe
even sooner because things are moving
at, uh, you know, at breakneck speed.
Uh, you are gonna see a lot
more content going through.
That's one of the things we are
developing capabilities of AI dubbing.
Number two is that in
Kevin Horek: many of them, sorry, can I,
can I interrupt you there for a second?
Um, the other, and how does that play?
Because AI, I've seen obviously generate
songs that sound like the actual
artist, when it's not the artist.
Is the technology gonna be there
with this dubbing that you talk
about where you could use AI
to talk in the original actor's
voice, but in a different language?
Yes.
Burhan Fatah: Because when you,
when you watch video, uh, on any
screen, video is essentially a two.
Separate files.
One is your actual video file
and other is your audio file.
Right?
And, uh, so you can process them
individually and yes, a, as a
matter of fact, those are some of
those, uh, fascinating applications.
Just think about this way, uh, let's
say you're watching, uh, one of the
older James Martin movies, okay?
And you want to replace the actor there,
uh, Roger Moore with Daniel Craig, right?
And you will be able to do
this kind of things also.
So now you can create your own
customized content in which you can
have your own favorite star playing
the role, uh, as opposed to Yeah.
Kevin Horek: That's interesting.
That, that's cool.
So, Okay, keep going.
Sorry, the second part I interrupted you.
Yeah.
Burhan Fatah: Uh, what
was the second part?
Okay.
Yeah.
In within our, so, you know, part of our,
part of our business model is, uh, free
content, uh, uh, which is ADSD supported.
And the problem is that in many
emerging markers, there's not
enough advertisement available,
so they don't even have any ads.
It to run.
So we are building tools which we are
gonna make available to the end user so
they can co and create their own ads.
And once the ads are created, then
we can run them in our platform.
So that would be a
great application of ai.
Number three application.
As we are building more data centers,
uh, you know, you have to find optimized
paths to, for routing your data.
And, uh, no one is better than
ai, automatic data routing.
Okay.
You know, Frankfurt data center.
Is slow.
So you run it through Lisbon or that
is slow, so you run it through Paris.
I mean, I'm just giving examples, right?
So that AI can do that.
And last but not least, which is
probably the most exciting part of AI
in video, is that AI is going to create.
Video, you'll be able to create entire
movies within two years based on a script
which you can generate through chat g bt.
You can plug it into an AI machine and
uh, uh, app and, uh, it'll be able to
convert that, uh, script into a video.
So the, those are the things,
uh, which are going to happen.
They are coming.
And we are very, very excited about it.
However, the next question, I mean,
in this space, uh, you asked about how
we are gonna be using AI and so on.
That's how we're using.
But in order to incorporate AI
into your data centers, uh, that's
a whole different challenge.
So what exactly you need, I don't know if
you want me to go a little bit into how.
That will be, uh, implemented.
Um, it's just a very
complicated structure.
Uh, first of all, you need to have
a cloud, but how do you convert
ai, uh, cloud into an AI cloud?
Um, it's a pretty complicated structure,
which goes several layers deep.
But let me just, uh, for the audience
for, uh, who may not be very technical,
uh, I just break it into three parts.
So your number one, your bottom part.
Is, uh, in this stack, which is the
AI stack, uh, in a cloud environment.
It, it includes all the infrastructure for
ai, such as your servers, uh, such as your
cards, which have AI chips built in the
video cards and so on, and video cards.
You may have heard about them.
They, they process a
lot of videos and so on.
So, uh, all other forms
of, uh, amend data.
So that is the base level structure.
So your capacity has to be, your
processors have to be very strong
GPUs, uh, rather than CPUs, although
CPUs will be applicable and all
the Ai, h hardware and so on.
So that is the infrastructure.
The top level are going to be the
applications, which majority of, uh, what.
An average person is gonna experience.
And those applications are the kind
of applications we are talking about.
I mean, dubbing content, uh, or
creating ad creation tools and so on.
There couldn't be all
kinds of application build.
Uh, I imagine there could be thousands,
if not tens of thousands of applications
are gonna be built, uh, on that.
And the third part is the middle part.
It's called the middleware.
And that's the most complicated part.
That's where everything happens.
That's where machine learning takes place.
That's where the operating platform runs.
That's where the data sets are placed.
Now, data sets are very large, uh,
databases of data, which can train
applications in specific areas.
If you want to train on video, you
need, uh, large scale video data sets.
If you need, uh, you know, training
in some manufacturing, then you need
large data of manufacturing where you
train these applications and so on.
So, Those are the three rough, uh,
I mean, areas we are working on.
The bottom two, which is the
infrastructure and the middleware and the
applications are gonna be, you know, some
will be ours, but eventually we are gonna
open up our data centers to third party
developers and so on who can come and
build their applications and our platform.
Fascinating.
Okay.
Kevin Horek: So.
In your example of, you know,
generating a movie, for example,
and I obviously will have the same
access, or everybody will have the same
access as some of the biggest movie
makers, you know, today and tomorrow.
What do you think is really
gonna separate, you know,
good content from bad content?
Because if we're all generating stuff,
that's pretty good or good even.
How is it gonna be different or, or
set us apart from something else?
Is it creativity or what do you
think is gonna be the differentiator
between kind of, you know, how people
survive in this space really, right.
Burhan Fatah: Uh, yes, very good question.
I think, uh, the days of
bigger studios will be done.
Okay.
And I think, uh, you know, mega studios,
um, I think, uh, most of, uh, the
studios are gonna become more like
high tech labs rather than studios.
You still need, what
do you mean by cabins?
What do that, sorry.
Yes.
Uh, so you still need, uh, talent
to write scripts and so on.
But, uh, I think the difference
between video created by an, uh, by,
by a user, I mean an average user
versus, uh, uh, professional studio,
the difference is gonna start getting,
is gonna get narrower and narrower.
Sure that makes sense.
And you'll be able to create, uh,
you still need a basic knowledge
and so on, which people will learn.
I mean, I look at YouTube and I
look at the early days of YouTube.
By the way, we were one of the very first,
uh, partners with YouTube back in 2004.
Five gone, then they launched, and uh,
YouTube has gone tremendous iterations.
I mean, it was really this.
Uh, herky jerky motion, people uploading
videos, and now you can see fairly
professionally generated video programs,
uh, by, you know, so I think it's
gonna speed up that process a lot more.
You will be able to generate, uh, fairly
professional looking content, uh, uh,
with the tools which are available online.
And you'll be able to do that?
Yes.
Kevin Horek: Interesting.
And then what do you
think is gonna separate I.
People apart then like
Burhan Fatah: creativity if we really
want to go really far into the future.
Okay.
And by that I don't mean a hundred years.
By that I mean, uh,
next four or five years.
That in terms of AI is a very far future.
Future.
Uh, I, I believe that.
At some point in the next, let's say
this decade, okay, you will be able to
create your own personalized content.
Uh, based on some of the content
the way you want to watch it.
You'll be able to plug in your own
favorite actors to what you are watching.
You'll be able to put
your own favorite voices.
You may be even to create more special
effects, uh, on, uh, whatever you want.
So if there are.
8 billion people, unless say, let's say 6
billion people, they watch video regularly
on the, there may be 6 billion channels.
Uh, everyone having their personal
preferences and the way they
want to watch and consume video.
And this means multiple things.
This also means that days of blockbuster
content, uh, uh, are numbered.
Uh, we have already seen that.
I mean, you just look at how many
blockbuster movies are produced today.
I mean, yeah, you have couple of top
guns and few, I mean, uh, there were
years when there were 10, 15, 20 mega a
hundred, $200 million movies were produced
because the user base is segmented.
So, No, we don't watch,
uh, in any one place.
Lot of content.
Content is already getting segmented.
The same television late night shows,
which used to draw a hundred million
viewers today, they draw 10 million
and they'll be lucky if they have 10
million viewers because we have so many
more options to consume video today.
That it's all fragmented,
uh, this whole space.
And this only means that the media which
is traditionally dominated by bigger
players is getting more Democrat, uh,
democratic media and in which, uh, even
average people will have the opportunity
to contribute and be very creative.
Kevin Horek: Fascinating.
We're kind of coming to the end
of the show, so is there anything
else that you want to cover?
To close out the show or any advice
to anybody that's in the video space
or looking to get into the video space
around kind of anything we talked about
Burhan Fatah: today?
Sure.
I mean, you know, this is, uh, this is
our goal is to, uh, reach out to, to very
large number of users, uh, conservatively
speaking, uh, couple of billion users.
That's our goal.
And, uh, this is the addressable market
we want to be dealing with, and we
like to capture over the next few years
at least 10% of that market, right?
So if it's 2 billion, uh, that will
translate to 200 million users.
Uh, it, it may turn out, uh, that
it can be done more than that.
So we are, uh, very
interested in talking to.
You know, various companies.
I mean, uh, by building this value
proposition together, we are looking for
partnerships, uh, with the technology
companies, with the content providers.
Uh, we are growing our user base
and we'd love to talk to partners
and even in investors because we
have opportunities, we'll be open to
speaking with investors as well who are
interested, uh, in, uh, in, uh, getting
involved with global markets because,
uh, That's the most exciting thing,
uh, we see as far as, uh, we can see.
Very
Kevin Horek: cool.
But sadly, we're outta time.
So how about we close with mentioning
where people can get more information
about yourself, the company,
and any other links you wanna
Burhan Fatah: mention.
Sure they can check my bio, although I
haven't updated it in a while on LinkedIn.
Uh, but uh, you can get some
corporate information on uh
si.com, which is a website.
Uh, you can go and watch
actually some free television.
Uh, sibu tv.com.
It's a web app.
You can, uh, watch there.
Similarly, we have for the US
market, uh, we have Android
and iOS apps in the app stores.
They are available as well.
Even though bulk of our focus is,
and most of our content currently is
targeted towards international market,
we still have, uh, enough offering for
the US audience to go and check us out.
Kevin Horek: Very cool.
And Sibu is spelled S I V
O o just for that correct.
Listening.
That's correct.
Well, Han, I really appreciate you taking
the time of your day to be on the show.
I look forward to keeping in touch with
you and have a good rest of your day
Burhan Fatah: man, actually
Tha thanks a lot Kevin.
Thank you.
Okay, bye.
Bye.
