Executive Interview
The Future of Data Intelligence Platforms with Adam Conway
Adam Conway · SVP Products · LinkedIn
all of a sudden they were just liberated and able to find this stuff on their own what about uh working with Ani as a product manager sequel when it gets complicated is really complicated they
should be able to leverage AI the same way as companies like Google that are AI companies and hey folks today we have an amazing guest we have Adam Conway Chief product officer uh data bricks we are super
happy to have have you with us and of course we have my favorite co-host Quinton hi folks and we want just to ask you like very basic question can you tell us more about your job and your main
responsibility yeah and and first of all just say I'll say thanks for having me Clinton and Yousef this is great I'm uh uh it's going to be fun to talk about this um yeah so my my job um you know
you know running kind of product management for a company is uh especially company like dat size is is pretty awesome um and I'll talk a little bit about that but I'll also say that
datab bricks is an unusual company in a good way um and that's that we have a Founder CEO who's really who is really the CPO which is actually like a really nice thing to have where I have this
person um who's my manager you know been here since the beginning has all the context has um you know a strong vision for the company who's um you know really has you know that kind of I would say
classic CPO role and as well as a CEO role and I'm I I'm I'm there to make him very successful um and also be a sounding board and all those things and manage the product management team so
it's a it's a pretty awesome setup and I think many successful companies have this setup where you know you look at you know you look at places like door Dash and Airbnb um and uh and Facebook
and other place or meta I'm sorry uh you know companies like this this is kind of the structure of the company where you have this you know remarkable founder who who really leads um leads kind of
all aspects of the company but is very product focused um so that's um I think that that's like one of the secret sauces of data Bricks now you asking know what is CPO so what is CPO it's
many things um you know of course managing the product management team is very important being involved in the company strategy really and product strategy making sure that that's done
properly um also making sure that sort of the apparatus of how do we um build products at data bricks is kind of working properly so you know are people talking to customers getting feedback
understanding what their needs are synthesizing that and then um and then working with engineering to make sure that that product gets delivered are they communicating it properly with the
rest of the organization so there kind of the structure that happens in product management especially at a product management team of our size um you really need to make sure that that gets
run properly and then finally there's of course hiring um you know product manager product managers are so critical to a company a good product manager can have a huge impact on on a product um
make their engineering team much more efficient um get the product out to Market faster all those kinds of things so a product manager has such a huge impact that hiring extremely well uh becomes
critical um and even one product manager that's that's struggling can have a really negative impact on the organization as well so making sure that you're getting people that can operate
within the datab brecks environment so so how is it working at the high level uh so we talk about the CEO being kind of the CPO so I assum is like sharing the vision that's where we
think the market is going that's what we should be doing and then you disle the information through the PM and the PM you know build the actual well design the actual implementation and it goes up
is it kind of an up and down cycle like that yeah and maybe um we can talk a little bit about kind of the difference between like tops what comes tops down what comes Bottoms Up um in an
organization like data brecks the you know there are certain things that must come from the team themselves you have to empower teams to be able to do to do their own research to have
their own hypotheses to validate those hypothesis to gather data to um kind of understand the road maps and build the road maps otherwise everything gets blocked on one person so you can't you
can't of course have have that and people don't like being you know micromanaged to that degree also so you got to give some AG but you know our company is we build a unified product which by the way a lot
of companies don't so I'll just that maybe I should stop there for a minute and say in Enterprise there are companies that build unified products and there's companies that don't um and if you look at like the cloud
providers that's an example of like a very distributed um setup like where you have general managers they make their own decisions they have hundreds of products those products may not work
well together they may work excellent together it may not work well together but there's no mandate that they work well together um and anybody that's used like one of the cloud products this is
like very obvious like um there may be some you know base level compatibility with things like like an Amazon and ec2 or you know storage compute security but that's really it that's really the the
limitation of the integration at a company like datab brecks it should feel like a single product it should feel like you're coming in and even though you might be using a data warehouse or
you might be doing data science or you might be using a vector database um it feels like a single product and a single coherent experience um uh and in order for that to occur you've
got to have like pretty strong tops down Mandate of like how unification is going to work You' really got to have a vision for how unification is going to happen and then you have to implement a
strategy to make sure that that cons that happens and also organize the org chart so that it happens I mean that's actually another complexity is if you separate the or chart and have like two
GMS that are competing with each other then they're going there's going to be a bifurcation in the product by the way this is called Conway's law not named after me um but every time somebody says
Conway's law at data brecks they stare at me um even though uh I'm not that Conway um but that's but that's like another piece is like you got to avoid the org chart complexity there as
well but so what is what happens we have tops down kind of we're a tops down kind of unification Vision strategy on how we achieve that the teams individually operate within that structure um and
build road maps um you know you know create potentially new products do those sorts of things within that environment and we just basically have very like lots of checks and balances I shouldn't
say checks balances that's not the right word we spend a lot of time kind of aligning on making sure that these new products these new capabilities work within that structure that's that's so I guess maybe
you said it right Quenton you like said as it go back and forth it does go back and forth um but uh um that's basically how that how we kind of structure structure the from what I've seen uh like working
at data breaks like I think people underestimate the complexity of having everything working together and building like a unified product and and I do think it's the main value of data bricks
because when you think about it like you can almost take every single feature of data bricks and you will find something doing it somewhere but like the value is really putting everything together and
just putting the governance on top of everything yeah and I yeah I totally get how how many how many PMS do we have roughly at dat bricks what's the team about 100 we about 100 PMS to you need a
100 pm to walk together yeah exactly and since you men like one of your mission is to higher pm like what skills are you expecting from from a PM yeah um so maybe I'll talk about like
I'll talk about skills but there's also things you know we also hire product managers out of college they have almost zero skills uh coming out of college we have an APM it's called an APM program
associate product management program so I'll talk first about like the core capabilities um you know to be a to be a product manager you I think there's like some pretty core like things that have
to be true about yourself you know it's it's a hard job you know it's it's it's a lot of work um uh so you've got to be able to like you know work work very hard um you've got to be smart be able
to think from first principles um if you don't the tech lead who's next to you and might might have like you know an IQ of 150 or might be a genius type person will like not respect your opinions in
that way so you've got to just like be very clear very very much able to kind of have those conversations you have to be very humble one of the challenges of product management is that even though
you know product managers paired up with a tech lead they work together and make decisions um really drive that product it's kind of on the product manager to make that relationship work um you know
product managers are often that connective tissue between the field between marketing between the customer and Engineering um so you've really got sort of be this person that's very
humble willing to listen to people but also you know sort of make your point um finally you got to be able to write uh and uh you know writing is really a critical capability because it's often
the way that you are you know you're sort of memorializing the opinions and the feedback from customers and all of this stuff and if you write very clearly very succinctly um that communicate
you're that's like the most important communication medium you have as a product manager so those are kind of core capabilities now becoming like an excellent product manager it's a very it's a long journey
you know you you've got to be able to do um uh you know because it's a there's so many skills that you need um and you you sort of have to be sort you're sort of like the the Renaissance person of you know
of of the company as a as a product manager because you've got to be able to you know work very closely with customers and interview them very well to extract like exactly what their
problems or what their jobs to be done are is what we often talk about now um uh you have to be able to do that you have to be able to synthesize that very clearly build hypotheses prove those hypotheses out
finally you have to be able to put that definition into the into the product you've got to have a lot of product sense about how do you release that product how do you introduce it to
Market you have to then Market it communicate it to the um rest of the organization you have to spend time um you know talking with the field talking with essay talking with um customers and
then do messaging all the marketing all of those sorts of things you also need to be pretty good at you know financial planning you've got to be pretty good at so you have to have all these like you
know there's many skilled product manager actually I'll say a product manager that's been doing it for a long time often can do the function of the neighboring teams like they can be a
pretty good product marketer um they can be a pretty good Financial modeler they can be a pretty good uh essay even you know they can like kind of pinch it for an essay Maybe not maybe maybe not quite
as good as most essay but uh you can pinch hit for one um you can you know help close a customer for for a certain deal um on their product so you have to have like all of those skills which is
uh I think I think that's the joy of product management right is that you're you get to constantly learn new new things and new ways of doing things I don't know yeah I feel like like that's a
lot would you say as a product manager you need to have like a strong opinion on things uh and really you know kind of knowing the market and have a feeling of where it's going or is it more like okay
like it doesn't really matter if I have an opinion because anyway I have to figure out from the customer and just you know get the information from the from the like the one was are going to
use the platform what you think I mean you have to have a a strong opinion but rooted in evidence what does that mean um you have to talk to customers all the time and not just like just have a chat
with them we have to sort of really understand their needs um there's there's this thing called the five wise it was um it was part of the Toyota you know um system that they had to forget what it's
called now but um you know where you basically have to keep on asking why to understand why a customer wants something because what a customer is going to say to you is you know I have
this problem right now it's like please fix it for me I need this button I need a knob I need a something but often times they that's that's the thing they're they're speaking to you and what they know is
possible and what they need right now but anything you anytime you start developing something it might take you a year to develop this thing in a year what's that need going to be you're also going to know like how
does is there un some underlying problem which you can just make go away like if they're asking for a knob maybe you can just eliminate the need for a knob for that Forever by changing the way you build
the platform so really understanding the problem behind the problem or what we you know what people call the jobs to be done is really a critical part of pulling information um kind of out of that out
of folks so that's a core part of this research so talking to customers is really really important um also looking at data is important um and I'll say it's kind of there's kind of a different
emphasis um there's you know product management is kind of split in two there's consumer PMs and there's Enterprise PMS I'm I'm an Enterprise PM an Enterprise PMS it really focuses
around many many many customer conversations um like kind of you should be doing you should be interviewing two customers a day pretty much every day for your entire career um that's a lot yeah um and that's kind of
the key and then doing some data because data is also is supporting part in consumer it's actually kind of flipped it's like mostly data with a few like um customer inputs but customer inputs are
are strange you have to do more um like panels you might have another organization that does user research you might have a those kinds of things so it's it's a different it's a little bit
more separated from the individual customers and the reason for that is that any and consumer any given customer is like one one billionth of your Revenue opportunity whereas um in an Enterprise
a large customer can be significant amount of Revenue you know you might only need a few hundred big customers to become a pretty big Enterprise and you can interview several hundred customers that's not
difficult I mean over the course of years and with multiple product managers you can have regular touch points with hundreds of customers um so it gives you a very good view into exactly what your customers
need and prospective customers by the way that's also important sometimes you need to talk to customers that aren't buying your product um so yeah yeah and I have like the $1 million question we
know that dà break say have built many many successful like features but what's the secret Source like behind building those those products and we can see an example it's like how Unity catalog
started and then it was integrated in so many features like having to do those cross meetings with so many like I don't know ml team future Star Team DLT and so on so what's the secret behind this
success you know it's like it's like all things it's like the secret is just like a lot of hard work um uh and a lot of good thinking um yeah so it's not it's not a secret so much it's just lot of
companies I think struggle they struggle to keep the discipline of doing this so let me speak about this kind of specifically um if you have a disciplined product management organization that has a
extremely strong like set of evidence around what customers need um not only that but you've actually done the other research to really know kind of what the jobs to be done behind those evidence are you now
have a very rich resource which you can use to make decisions secondly you can look at um kind of your you know what problems have to be solved to kind of large in the market maybe people aren't
asking for something like Unity catalog which by way nobody was asking for Unity catalog they were asking for catalog improvements but they were not asking for Unity catalog they're asking for
governance like I need row based uh or I need um you know um I need what's called you know column and row based you know controls on on the data um they need a very specific set of things
and we had a lot of research on all of these things it was really sort of uh I think taking all of that feedback and be able to synthesize it into something very special kind of required
somebody to really understand where the market is going and by way we had some pretty amazing people working on that you know people like um you know M worked on it um uh himself um one of our
you know um most senior kind of distinguished Engineers is working on it um one of our most senior product leaders was working on it and kind of together you know they really spent a
lot of time figuring out where the market was going understanding what was needed um and then I think also taking advantage of the advant of the inherent advantages that datab breaks had data
bricks had some core advantages when before Unity catalog was built one was we did both data and AI at the same time so we had this very clear we understood that this problem spanned multiple kinds
of workloads and then the second thing was is we had this we were data L only uh which was kind of an unsolved problem at that time like how do you govern files and tables in the same way
so it was just these very basic building blocks that we had because of the growth of the company where we were able to we had sort of an an advantage in our thinking and our product that kind of
led to us building that and and by way I think you know we're talking about Unity catalog many many people listening to this might not understand the importance of unity catalog Unity catalog
is the even though it's not like we don't like charge for it independently it is you know it's part of the produ it's part of the platform itself it is over and over again customers tell me
it's the our best product like it's the best thing we've ever done um it's um it just simplifies so many different uh ways you do things and basically it provides a central place to find govern
manage all of your data assets all up and that means like AI models to um tables to files to environments to other things that are out there um it's the core of how you do cicd on data breaks
it's the core of um you know um all of the AI capabilities and actually probably maybe the most important piece is all the AI capabilities we have utilize Unity catalog as kind of a
backing store uh to to evaluate things so all right that's a long that's a long answer on and a Meandering answer but uh hopefully that's cover everything here so you mentioned you know all the like
customer meetings to get the feedback and and asking many questions really understand what they need uh but once you have this information and maybe you think about a product you should be
building or maybe you know improving something adding adding features how do you go into deciding what to put as a priority and what to build first to you because I guess you have a balance of
maybe potentially existing big customer and you can kind of track Revenue out of it but you would also have potentially features you want to do for like the market you want to concure so how do you
balance that yeah I mean I think if you want to be a big if you want to be a big successful company you have to think long term all the time and at the risk because I guess if you have like a big customer they ask
you something and you're like hey no you know what we won't do it because we don't think whatever the reason that's what we want to do oosh yeah I mean that's um yes like of course you try and meet the needs of big
customers but many companies get trapped in this environment where they're constantly building features for big customers and they sort of never get so you sort of never get real product
Market fit where the product takes off on its own now you're kind of stuck in this mode of being you know sort of building custom products for all of your big customers and and I think that's a trap
that that many Enterprise companies find themselves in if you want to build like an amazing product for the future it's not going to come because you have big customers asking you for um you know a bunch of security
features or a bunch of things like that you're going to become a great company of the longterm because you build phenomenal products that people love um that are consistent across all customers
um you know uh and I think that's that's like the most important thing so if you're constantly thinking long term about what you're building it allows you to do that now of course you carve out
some time for a short term like like I need to like solve you know a set of problems of course you do um but I think over time if you'll regret some of those short-term things like maybe customer
can wait six more months and you can do it right they'll be happier and the rest of your customers will be happier and but I think you know we've all seen this at data breaks customer
needs something instead of doing this quick expedient thing we do the longterm Thing by unity catalog was another example of this um you know some of the asks we had around you know U things like column
masking and things like that um you know it took us ended up taking us years to to respond to that but when we did it we did it so well that customers just adore um data bricks for it um and
that's that's I think I think it's just a better way of of operating um you know trying it's much better to build workarounds by it's much better to have like a phenomenal field engineering team
that can also kind of help bu those workarounds within those customers um so that they can sort of and by they they they then bear the pain of these customers but if they can like solve the problem without having to
affect product and gives us you know some time to actually do it properly uh in the product that's just amazing that sort of gets you that allows you to sort of live in this world of building the
future rather than again sort of like that problem of like dealing with current needs which basically means you're dealing with the past by the time you get it done so it allows you to you
know like I said build innovate in a way that you wouldn't otherwise be able to and I remember like um when I first heard about datab brace back in 2017 we had this like title like making
big data simple like since since like for the last three years I keep hearing like uh making the product simple accessible and I want to know like what's the definition of Syle uh in a product
development because I know it's like yeah I'm building a simple product but what's the key uh of having a like this uh success behind this simple like because I know it's like short word but
it means a lot behind yeah um you know I think it's sometimes Simplicity is like a very it's it's like a very difficult to understand term uh and I think sometimes Simplicity means not capable
sometime which but that's not good you don't want that end of it especially in a product like data brex um it can mean you know a lot of things so I think you know it should be like in a
product like data brex which is you know where we serve you know data practitioners sometimes people are you know pretty hardcore data practitioners you know writing code building data
pipelines that kind of stuff all the way up to business users who do not understand SQL you know basically their knowledge sort of ends at a spreadsheet like how do you how do you serve kind of
that class of people those that huge that huge group of people and I'll say there's a few things but there's a few things that are Universal even when you're looking at that kind of range of
problem which is basically the problem that data RX is trying to solve and that is the fewer Concepts you have the better off you are so and by this is the beauty of unification is that the concepts get
dramatically reduced and and I'll say that unification is extremely hard but this is where this is where the value really pays off it's having one place where you do governance which we keep on talking
about governance but governance is often the area that adds so much complexity and actually it's probably the greatest complexity generator of any um of any system out there having one set of
Concepts there means means that people if you learn if you start you know writing SQL and you're using the data warehouse and you're having a you're seeing how things are working and then
you decide okay I'm going to go write a python notebook and everything is the same all the concepts are identical except for the python itself then it makes it it dramatically simplifies the platform so one is kind
of reducing Concepts and by way this also means knobs and things like this you know like different reduce the number of settings most set are just not necessary if you really think about you
know um except for dark mode everybody knows that's like a most important setting you have to have in a product um um but uh uh the um but you know if you can reduce those number of settings
down to near zero then you're in a much better State um and by that's hard in the Enterprise because what customers want is they want more settings but they all want different settings so then
you'll end up with like this huge proliferation of settings so you have to so that's like a very hard uh condensing problem second thing which by the way I don't think people think about a lot is
standards you're allowed to do things that are hard if it's just standard across the industry um uh and I'll give you a great example of this which is spark everybody knows everybody that
does data at scale knows what spark is if you're doing data processing you know you know py spark or you know um the Scola API or you know um spark SQL or you know or sparkly are you know you
know one of these apis those apis are not simple they can do you know pisar can do many many things you can it's a very very powerful system but it's a standard so if you you learn it you know you learn it in
school um so if you adopt standards um or build even better yet build standards they get adopted by the whole industry then you know then that's Simplicity I mean that's that means that
everybody is willing to learn it and use it um so I think those are the ways that I think we think of Simplicity it's like unifying Concepts so reduce Concepts reduce settings as much as possible
which is always a hard challenge uh and then finally it's like adopt standards um and make that work well I make maybe make another comment you have to use the product all the time your your product management
team your sa team your engineering team you got to just like build systems within the company to make people use the product over and over and over and again and then measure how they use it
and make sure it gets better over time um we do this through something called CJs um but companies have many different mechanisms for this but it's kind of this idea of like you got to dog food
the product you got to use it every day got to make sure that that happens and then record all those issues that people have and work on uh and maybe I should have mentioned that first because that's maybe the most
important thing but that's like that's like a pretty awesome um kind of mechanism within a company that makes things just much much easier to use yeah and I feel like measuring like it's really hard for me
to measure for example we do demos it's really hard for me to measure if the demo is simple or not simple because once you do it and you know everything about the product sure know how it's
working um but like having someone that you take you know being brand new in the like whatever you do and and making him try see if it's working see if it's simple I think I think that's a really
good way to measure Simplicity yeah yeah that's say that's like a little that's good that you brought that up there's a there's this trick that we pull a data breaks where every new
employee assuming you know uh every new employee has to basically go through and use the product and achieve some goal in the product and then provide feedback to us that's like they we one task um
and you know I'll say it's funny one of these funny things where sometimes new employees get addicted and they'll like say oh that was fun and they ask for another one and they'll do like six of
these like Journeys across the product uh in their first like few weeks and then they'll know everything about how data bricks works and they'll know they know where the good is they know where
the bad is they know where the problems are they'll feel motivated to improve them they become very happy employees U uh we we recently hired a new director in ux um and and um and she never um she'd
like never been in the data platform space you know that was like not her I mean she understood data she had some basic you know understanding of data and she I think she did five of these in her
first few weeks at data brecks and was just like addicted um and that was great and that's like an amazing um amazing outcome for a for a project like that so all right so I wanted to do like a small
segue and and talk a bit about about AI because AI is everywhere nowadays um so I guess we can start by talking about um like because dat brick is doing a lot of AI product um AI is
moving so fast right now like if I were like a PM on AI I would be super scared because like whatever you want to do this year might be completely outdated but yet you release the product it's
going to be like you start to have some adoption who knows if it's going to be what you actually have to build in a few years but at the same time you don't want to miss the Innovation train so you
want to make sure you do build something it seems like almost like super hard to balance how do you do that how do you like do you do you do you wait a little bit and just make sure you go to
whatever you think is going to stay like for example if we take rag rag I guess used to be a big thing last year and now everybody's talking about agent and agenting stuff so if you built the rag
stuff last year well you kind of missed on the the big new thing so H how do you do that yeah I mean I think there's a few things but I mean I think one of the things that I think one of the luxuries that we
have at data brex maybe that most companies don't have is that you can be part of building the future if you have a good AI research team um so I think that's one of the benefits that we have
here at data bricks um that that most honestly just most companies even companies of our size don't really have um uh and I think that's allowed us and I think the fact that we've been doing AI since day one
of the company which is I think is very rare you know for to be at a 11-year-old company that's been doing AI since the very beginning um uh I think that makes us unusual so we've always had this
DNA so that's I think sort of like a I should just point out that we have maybe maybe we have like a secular advantage in that way but I think you have to be willing to just build and throw something out and build
something and throw something out and sort of like keep on executing um delivering capability delivering value as quickly as you can make sure you're actually delivering value make sure customers are using it
that they're willing to adopt it uh and then and then move on and the beauty is is a lot of the Core Concepts don't change that's the beauty of of it is like even when if you look back a couple
years um you know rag has existed for most of that time um so Vector database is still part of an agentic system um uh so we have a very very good Vector database um capability um you
know uh being prompting has like grown dramatically and things like DSP and all these capabilities and the ability to to Think Through to build those sorts of systems that's had value over this whole
time so some of these things the beauty of is some of these things are changing very rapidly but some of them are staying the same and some of the Innovations you came up with a year ago
still apply others not so much I mean certain things matter less now um uh you know and that's so I think that's where um you just have to be willing to move fast and throw stuff away if it's it's
not if it's not working anymore um and it's a huge investment but like I said it's a big business for us so it's worth it so I guess there's no trick there there's I wish there was a trick uh but
I think it's you know like like most things you know overnight success requires many many years of hard work um so and and I feel like like no but what you say makees sense for example when
you take the like the evaluation Parts something you will need like even if the the way we build the AI is going to change like maybe it's going to be new model new something you will still need
to have some kind of eval to know if it's working not working double done if if we double done also on things like that that you know is going to stay forever it's also like a good
way to add really good value um oh yeah while having that for a long term yeah F thank you thank you that's a also a very good answer I should have said that uh yeah so a lot of the a lot of the
tools and systems that you need um to build these systems have actually remained constant Concepts like guardrails and Concepts like valuation and Concepts like um a lot of what we call agents today
was called you know stuff like Lang chain and stuff like that before I mean it's a lot of those things are you know similar Concepts and similar scaffolding built around um the models and other capabilities that that
carries forward so yes we as the names have changed and the terms have changed a lot of the infrastructure is the same fortunately and and and so and so we talk about AI in the product what about uh working
with an AI as a product manager like do you I don't know do you use REI when you do product management other than I would say the basic stuff like you know summary of meetings and things like that
um yes I do I do use AI um we do and that's b a good example is yeah summaries of meetings things like that are very very important um also research is is very nice um you know um things
like internally we use things like we use Rag and we use pabilities like that um by it's also a very good way of communicating um with uh you know the the field engineering team as an example
like now they they they can use utilize systems uh inside of data bricks to find out hey when do X feature ship you know you can go ask that question so that also helps on the communication side I
think there's some things that have not been cracked yet by um uh by it um by gen and I think the first thing that I've seen is I've not seen the kind of writing you do in PM where you write something very suin very
to the point um you know very you know it's very pedagogical it's not quite there from a gen from gen yet so you can't if you're P if you're trying to write a you can't have gen write peties
for you you know you can't which is a requirements document you can't have gen really perfectly capture like what is the most important nugget out of a meeting um but I hope we get there at
some point um but we're not we're not there yet so I think those bun of those aspects are not are not quite there um but every year we add new tools to our tool belt with AI and it's been awesome
actually we have a project right now um ta trying to um kind of pull together more knowledge uh about our customers um uh in a single in a single place so that we can ask the question of like you know
how important is X feature or who's who cares about this this kind of this kind of problem um uh you know those kinds of those kinds of questions and then also help us figure out like which customers
are the best customers to talk to about a specific new feature or things like that and and so what about even getting feedback at scale using gen like you mentioned doing for example two
interview per day but like it's super time consuming could could you imagine I mean I don't know is the industry doing some kind of you know massive survey through gen to get some some feedback
from the customer I do I mean I think there there is some advancements in that area but I think the core work of product management which is just sitting down with the customer and understanding
their world empathizing for what their needs are um sort of I think the key thing and maybe I in most Enterprise companies like data brex my job as a product manager is different than the customer job that I'm
talking to um they don't have the same role as me um I work I work in an R&D department they may or may not work in an R&D Department um I have kind of no operational responsibility for like a
data Pipeline and they that might be their main job is like how is this data pipeline running I don't want to get a page at two o'clock in the morning you know how do I make sure that the data is
reliable that's not we're we're just live very different lives and I need to understand their life and understand what their day looks like understand what how what they go through
um and there's not a clear I I don't know there's a clear way to get that even out of a survey or out of um out of gen uh particularly you sort of have to sit sit down you know face to face via Zoom or
or face to face in reality and really understand them ask many questions probe deeper and deeper see how they work you know understand what they're doing and I think that's there's not a gen replacement for
that yet that I've seen yeah no no that's a good point because I didn't mind only the like the getting information U but you're actually also learning from the customer and I guess
you you need that discussion to have the learning the learning steps that makes sense yeah also like how it's phrased because I can say the same sentence but depending how I pronounce it it may be
different like whether I'm angry or maybe I'm happy or like those sentiments I think still the like agents cannot understand them yeah per they can't do it perfectly yeah exactly but I
think we get more I think we've gotten a lot more efficient product because of gen that's already happening you know I utilize j i 100 times a day maybe um to do to do various things in my job or whatever it ends up
being and that's that's made me much more efficient all right so yeah I feel like I wanted to ask at the end you know if you think that's product management excuse me I wanted to ask at the end if product
manager is going to be the last job a will replace but I guess we have the answer I think so I think well I think one thing about J the last thing j is going to replace is anything you do with your hands
right and I know yeah wait wait for the robots to come but yeah yeah you know but I'll say I'm a mechanical engineer uh undergrad that was what I studied and I was a mechanical engineer for a few
years and there's no gen tools there there is generative design it does exist and it actually is very very cool but it requires you to set all these very rigorous constraints and Advance um and
then what it produces might not be manufacturable so there's like this so I think like we're living in this world when you're dealing with code and you're dealing with text ji is amazing right now we have
like another like set of evolution before it's can do perception and you know understand more complicated you know non you know I guess the beauty of of text is there's quantum there's these tokens
they are like there's only so many of them in the world and so it's significantly shrinks the world's complexity but in the world of um most of these other things is not that so there's a lot of innovation coming in
that area and I think we're very all very excited about it and I think that's an those are probably the last there he has to be um I actually I won't say the word disruptive because I don't like
that word I don't think you know that's kind of like the Doomsday discussion that everybody has but I think it's it's they're the least augmented jobs today like they're not nobody's making your job better I mean I
don't think I think we're a long ways off from having a product manager that can be an AI product manager just like we're a long ways off from um you know having like a perfect AI developer what we're seeing right now is
all these like amazing tools we're becoming like the B bionic man or woman in developing code you have all these augmentations that are making you substantially better you know things
like um you know cursor is amazing but our own like our own assistant in the product is like shockingly successful and customers I'll hop on customer calls and the customer will be upset about
something maybe we SLI a date or something like that and then at the end of the call they'll say but assistant changed my life you know they'll tell they'll tell me something like that
which is um and you know pretty much every customer that can use assistant is using assistant right now it's like the it's almost like 100% um so it's a pretty amazing pretty amazing like thing
that's happening with this kind of augmentation just number of times I was forgetting like how like structure of an API instead of going like to open the browser just go to the assistant say hey
I need to do this and then you have like this and you have the answer quickly oh yeah I I the same thing or like window functions who knows how window function I mean like by the way uh because you
mentioned like uh you know being simple you have to be the nor like I don't think squel SQL is simple at all I just think everybody know sequels it looks simple but I it's a really good example
to to the points you were making before no I but I agree with that as well I I can write Python and like write do all this complicated stuff but as soon as a SQL query gets complicated it like
something in my mind breaks having the ability to have gen just generate that for me is amazing but I think you know we haven't talked about and we haven't talked about our product very
much but maybe it's kind of worthwhile just talk a second like we this we think the future is this thing called Data intelligence um uh in our space and data intelligence is really
where geni you know democratizes access to data because today you have to know python you have to know SQL we all we just talked about how SQL when it gets complicated is really complicated
we see this as like being very core to like democratizing access to um uh to data but also we also see data intelligence as saying that's also going to democratize access um the the ability for companies
to build AI um you know we want you know we don't you know our customers are large Enterprises um their main business is not delivering AI systems but they should be able to leverage AI the same
way as companies like Google that are AI companies and um be able to kind of operate in that level and I think we see data intelligence as a way for us to get there and I think that's like and I
think so so in our world I think this is what we see like what's being what tools are we getting data intelligence is really this this thing that we're that we're getting the benefit of
and we're putting into our products and I think aibi is the the like one of the best examples we can give like you can like I'm not a data analyst I don't know how to build dashboards but I just open like uh the
dashboard and start like asking like putting some prompt like Revenue versus sales and I have this chart built automatically for me do the same thing for all the other charts and then if I'm
still my dashboard still not good I can just open G said okay what tell me who is the best sales like in my company and then I have like in of having to write this complex SQL queries as Quin
mentioned you have like just text hey youf is the best sales I'm I'm not I'm kidding I'm not the best but you have this answer like done for you and this is like like the best example you can
give with the uh aibi or this data intelligence yeah I mean it's um just I mean just within data brecks you know um we have this funny thing marketing uses a you know uses aibi all the time
and they're so excited about it because they're all of a sudden have access to all this data they never had access to before they don't have to sense anything off to an analyst they don't have to get
a dashboard built that that that you know there was a period of time where that's all they wanted to Market because they were so excited themselves about it so it's like this um which is which is
pretty funny um uh but I think also just shows the power of like this is a team that had to rely on data analysts in order to get their to get their results and all of a sudden they were just liberated and able to
find this stuff on their own um and it's and it's just it was is really kind of a a night and day change for that organization and I think we view that as like this is that this is just the first
example of just many many of these um and imagine just how more how much more effective you can be if you don't have to round trip these questions through somebody through a human that has to
like schedule the time and find the time and may give you the wrong answer may not ask the question maybe ask the question wrong the first time I mean I don't know that's I think this is
amazing and I and I think if I can summarize all this session first if you are considering being a product manager in the future I think this is a recording this this session this is the
one you should be watching like having some some advices and also if you want to apply to data breaks you have some in you have some insights like how to prepare your interview and I truly
believe like the last part would be more about aibi and believe Chia and Miranda are going to be happy because we promoted dashboards and gen at the same time yeah of course of course well but also I think I mean the
beauty of it is is that it's infused throughout beauty of unification is you don't just get data intelligence inside of aii but you get it throughout the entire product again this is by going
back to unification is we've looked at every single place where we can make the product dramatically better with AI um and of course you don't see it anywhere obvious as you see in aibi because it's
you know literally is um uh the the interface that you would access as a as a maybe a non-technical user to use to get data so thank you so much Adam it was super exciting Cole and thank you for
your your insights and uh looking forward to having you back again for yeah maybe maybe next time we'll we'll talk more about the product uh and sounds good for sure sounds good well
thank time with you youf and Quinton forward to look forward to next time thank you