馃攳
Databricks CEO and Co-Founder Ali Ghodsi on Sky News - YouTube
Channel: unknown
[0]
well as stem was just saying data is
[2]
more valuable than ever and that was
[4]
also recently shown starkly by the 28
[6]
billion
[7]
evaluation of the machine learning and
[9]
data engineering company databricks
[11]
well the us firm's data platform is used
[13]
by companies like
[14]
h m nationwide sky's parent company
[17]
comcast
[18]
and 40 of the fortune 500 it's just
[21]
partnered with google cloud to expand
[22]
its global reach
[23]
and i spoke a little earlier to his
[25]
chief executive ali godzi
[27]
it turned out that all those things that
[30]
have been around many many decades
[32]
they actually didn't work very well but
[34]
around 2010 or 12
[36]
people figured out if they use much much
[38]
more data on modern
[39]
computers you get fantastic superhuman
[42]
results
[43]
now central to what you do is something
[45]
called lake house
[46]
architecture for the uninitiated what is
[49]
that
[51]
well it means people are already storing
[53]
all their data
[54]
in the cloud in something called data
[56]
lakes but they can't actually make use
[58]
of it there if they want to do machine
[59]
learning on it they have to move it
[60]
somewhere else
[61]
in these things that are called data
[62]
warehouses so we've cracked the code
[65]
on enabling these enterprises do machine
[68]
learning ai
[69]
right there on the data lake so they
[71]
don't need to necessarily move it to a
[72]
data warehouse
[73]
therefore we call it lake house so
[75]
portmanteau of those two things
[77]
right and what are the sectors where
[79]
you're seeing the most growth
[81]
in demand right now you know
[84]
that's a great question because when we
[85]
started the company i was trying to
[86]
figure out what's the sector we should
[88]
focus on
[89]
and i was always surprised it turns out
[91]
this stuff is applicable in every
[93]
vertical in every sector
[95]
you know we have people in healthcare we
[97]
have people in
[98]
mass media we have customers in you know
[102]
retail we have customers in the public
[104]
sector we have
[106]
you know customers that are in the iot
[108]
you know equipment space
[109]
so this technology is applicable
[111]
everywhere i think you
[113]
i'm you're i'm right in saying you've
[114]
done some very interesting things in the
[116]
healthcare sector in particular tell me
[118]
about those
[120]
yeah so you know a lot of the vaccine
[122]
development that's been going on a lot
[123]
of the pandemic fighting
[125]
behind the scenes they use data breaks
[127]
and they use ai
[128]
in particular to look at the genomics
[130]
and find genes
[132]
and gene markers that are responsible
[134]
for you know diseases and you know
[136]
phenotypes as they call them and this
[139]
has taken away a lot of
[141]
sort of man-hours physical labor that
[143]
previously wouldn't have just made this
[144]
work
[145]
impossible to do yeah in the genomic
[148]
space
[148]
it would probably be hard to do at all
[150]
if you didn't have the machine learning
[152]
uh you know because
[152]
this is huge dna sequences and you're
[155]
looking for that needle in the haystack
[157]
that potentially is responsible for you
[159]
know some disease so
[161]
it wouldn't be even possible with this
[163]
now you've just raised a billion dollars
[165]
from
[165]
investors how are you going to be
[166]
putting that to work
[169]
yeah so you know expanding
[170]
internationally is very costly it's
[172]
almost like starting the company all
[174]
over again you know every country
[175]
different legal code
[177]
different partners different hr so we're
[179]
gonna aggressively
[180]
expand in europe emea
[184]
in asia asia pac and also in latin
[186]
america so a lot of it will go to
[188]
international expansion
[189]
but we're also investing a lot of it
[191]
back into r d so our
[193]
percentage of r d spend as total
[195]
companies spend is actually increasing
[197]
year over year
[198]
so a lot of it is just continuing
[200]
investing in research and development
[202]
and the fact that you are investing so
[204]
much outside the united states
[206]
is that because the valley has become
[207]
too costly a place to do business or is
[209]
it just you're going where the customers
[211]
are
[213]
no i mean look europe is a huge market
[215]
asia is a massive market you know cloud
[217]
adoption
[218]
is you know sometimes bigger in some of
[220]
those places so
[221]
you know we're just going where those
[222]
are us is not the only place you know
[224]
where
[225]
people do business and some of your
[227]
shareholders it's some of the biggest
[229]
names in tech the likes of alphabet
[231]
amazon microsoft salesforce i mean some
[233]
of these are your customers as well as
[235]
your shareholders
[236]
how do you manage those two separate
[239]
relationships
[241]
yeah so look this is very early days in
[243]
the ai market
[245]
we're super excited uh you know these
[247]
cloud vendors they're now seeing that
[249]
all this data is moving from these
[251]
on-premises deployments into the cloud
[253]
and we're a killer app
[255]
for those cloud data lakes where all the
[257]
data resides
[258]
so it makes a lot of sense for them to
[259]
invest in us you know we drive lots of
[262]
consumption we launch
[263]
five million machines every day in the
[266]
cloud
[267]
uh but you know they're also customers
[269]
they also use our framework so you know
[271]
uh we try to be as you know customer
[273]
obsessed as
[274]
possible and follow whatever our
[276]
customers need and you also
[278]
obviously have a lot of traditional
[279]
financial investors the likes of
[281]
blackrock flank
[282]
franklin templeton and so forth are
[284]
there interests
[285]
in you and their demands on you
[286]
different from some of those
[288]
trade investors yeah i mean i think the
[293]
these firms that you named they are very
[295]
long-term
[296]
right and you know frankly speaking when
[298]
they invest in this round
[299]
we took one billion investment but the
[301]
truth is they wanted to deploy probably
[303]
three four billion dollars
[304]
now we didn't want to dilute that much
[305]
so we you know shaved it down
[307]
but they're thinking the long game and i
[309]
think these kind of investors they're
[311]
thinking you know
[312]
ai machine learning ai is going to be a
[314]
huge market
[315]
over the next 10 you know 20 years let's
[318]
get in now and let's just invest and buy
[319]
more and more and more if database stock
[321]
goes up we'll buy more if database stock
[322]
goes down we buy more i think that's
[324]
different from maybe
[325]
if you're more short-sighted that's an
[327]
astonishing uh thing you've just said
[329]
that you could have raised even more
[330]
money than
[331]
you actually did i mean presumably this
[333]
means you're in no rush to
[335]
go to the public markets well you know i
[338]
get a lot of questions around that
[340]
we're gonna assess that you know when
[341]
the timing is right but yeah we're not
[343]
in a hurry
[344]
um you know you can get a lot of these
[345]
benefits of ipo and already in the
[347]
private markets today
[348]
when the time is right we'll also
[350]
execute an ipo
Most Recent Videos:
You can go back to the homepage right here: Homepage