there has been a huge search in the data growth in almost all companies right big medium small everywhere and if you look at it like 20 years back even 1 Megabyte of data used to be huge we used to think it's a huge data but now it has gone to stuff like paby exabyte and all and what not we will come later we have to see early 2010 um is when I started hearing May in 2012 13 more on um the whole digital transformation initiative right and that actually brought in a lot of focus on
um eventually CIO side and then you know from there moving on to uh establishing or creating a brand new role CDO so slowly this Chief data officer role has moved more from an IT perspective in terms of just data management to also a business part of it in terms of seeing okay how can we help business uh Executives generate value using the data let's just say cxos can make informed or smart decisions right for business growth which means the r CDO actually need to have both side of um understanding background or maybe even work experience
right yes yes [Music] hello and welcome to AI connect um it is wonderful to be back again with a with another episode and in today's episode we are going to talk about a very uh new brand new topic that we haven't touched upon so far and that topic is building a worldclass data organization so this is going to be a three-part series the context of this discussion today is in today's digital economy data is in just an asset it's the life blood for Enterprises to make decisions and drive Innovations as organization capture unprecedented volumes of
data at this time the challenge is to organize that data analyze that data and derive value from data that drives important decisions that drives business growth yet many organizations are fundamentally struggling with one question considering that Enterprises capture and generate data across various touch points this unprecedented volume of data is sitting at their base the challenge is that you know the the central question is that how do I use this data to drive or derive business intelligence that drive my business growth this challenge has given rise to a new organizational imperative the need for dedicated
data centers of excellence led by season data leaders who can bridge the gap between technology capabilities and business outcomes in this first episode of our three-part Series in AI connect I'm going to invite anirban maunder who is the co-founder and CEO of pre decision solutions to discuss this topic in depth and we will go deep into this topic to unearth some of the important foundational need and the drivers for building a worldclass data organizations so with that Anan welcome to AI connect thanks J it's nice to be back on this absolutely and I think today
we are uh here with a with a very important topic which is timely and which is at the center of U almost all Enterprises that is data right so with a huge jump in or The Surge of generation of data and capturing data um do you think how do you think organization is taking care of that right is that all under their control uh I think it's a good question uh whether everything is under control uh let's see but yeah I mean there has been a huge search in the data growth in almost all companies
right big medium small everywhere and if you look at it like 20 years back even 1 Megabyte of data dat used to be huge we used to think it's a huge data but now it has gone to stuff like PAB exabyte and all and what not we will come later we have to see and again there's a variety of data so each organizations especially the bigger ones which are very complex organizations they have so many different you know software systems you have your Erp system your some hrms or CRM and then lots of small small
you know pockets of applications everyone is uh taking data and everyone is generating data so that creates a huge amount of data which will kind be transactional data whether it's Financial or sales or purchase or all these kind of things you will have customer data employee data operational data then a lot of people have the sensor data also iot and all have become very prominent now and that also while each device may be generating very small amounts when you have a huge amount of uh or huge number of such sensors that also leads to a
huge amount of data amazing and then you have the unstructured data you have all kinds of documents you have emails you have you know customer uh call records audio records and their transcripts you have images videos and whatnot so it becomes very difficult unless you have a set of you know processes and guidelines to manage this data and every every day you are generating it's not like you're stopping generating data even on a day on a Sunday or any other holiday when things are not working even then you're generating data always on always on right
and then you on the other side you have people who are consuming this data they need this data for various purposes whether to have the processes going on or whether to decide something whether to do something so then are consuming the data so one part is generating other part is right and these are internal data you also might have external data coming in right so there's a huge data growth and other thing is that also the consumption has been uh LED because today if you see in many companies most of the folks have access to
the data because they have laptops or mobile devices or some other device to connect to whatever data is relevant for her you know activities in the company so which was not the case 20 years back when people so many people may not have access to you know systems so if you look at even a retail store and the store individual store so many people have access to uh different know data terminals so all this is causing uh lot of churn and the fact the companies what they're seeing is that their data infrastructure is increasing in
a very rapid manner so they are storing data so earlier what we didn't have the cloud 20 years back it was primarily uh hard disk and then you used to take it off to tape drives right the data will be stored in tapes so even then you had the regulations which will say such data you store for 5 years or 7 years or 10 years which are still there but right now uh you use tapes only for long-term archival you're not looking at it to uh take out data immediately if you feel that you're going
to need it over the next 2 months or 6 months whatever so that's where the cloud infrastructure which has come in as helped a lot but it also led to a sprawl data sprawl as I could have to call it and there you have uh data all over the place correct and that leads to data management challenges you don't know what is the data coming in how good or bad that data is in terms of the quality in terms of the consistency in terms of the completeness so how do you manage this data so that
the users of this data are getting the right data which based on which they can take actions Bas based on which they can decide certain things and the other part that also comes in that all this data that has come in uh has led to lot of new technologies so you also need the talent to manage you know this data and then do analytics and visualization and you run data science models on that so unless you have the right side of talent you cannot have any competitive advantage over your rivals absolutely so um keeping that
in the context right and one in one side you have this whole um piles of data that are you know being generated 24 hours 7 Days 365 days um on the other hand um organizations are faced with um what are those data that are really useful what are those data that probably have a junk that means this whole data management piece um and then um with the charter of um getting the right data putting them into the structure which obviously means a lot of work right and um keeping in also mind that these data are
also created by different organizational functions right right um some of them are external facing some of them are internal facing so clearly this isn't um a charter for everybody but you know at the same time they are responsible or they are generating data so someone in the organization obviously is um is is need basically needs to step up and take the Mandate which is why this rise of different roles Like Chief data officers and you know other important roles came up right so um what I mean can you give a little bit of context of
how these role actually have um come about in the recent past and uh and give a a little bit of um or share a little bit of um context in terms of how important these roles are becoming in terms of organizations overall um not just the data management which is the internal part but um you know in a way kind of become a data Steward to contribute towards organizational growth by bringing in those insights uh or maybe uh intelligence that is actionable sure yeah so uh if you really look at it uh earlier or even
now in lot of companies the whole data piece uh in terms of storing and then Distributing data to the users is under the cio's organization the ID team and that's where it was even earlier days so the IT team may set up a data warehouse you know based on the Technologies available in those days and create that from which some analysis may be done someone will take out but slowly they realized that it needs some amount of effort in terms of data management data governance to ensure that the data is right and the initial push
came more from Regulatory and compliance perspective than anything else so one of the key acts that actually pushed it this in the US was the servant of SLE which needed you to do lot of stuff in comply with lot of things in the uh Financial Services space right and then of course you had similar things in Pharma in healthcare and all so we there to submit lot of reports and then get a lot of things audited not just the financial part but also in terms of your uh drug trials or any other uh compliance things
so there you needed to have the right data to and then crunch it and put it in a report in the Raw dat as well as the whatever calculations you have to do so that's where the people realized that we needed someone to look at it and play that role so that's why the initial Chief data officer kind of role started coming up the actual first Chief data officer is said is that came in Capital One in 2002 okay which is like almost know 23 years back and surprisingly the next company to announce such a
role was 3 years later Yahoo in 2005 so it took that much time even in 2010 there almost just about 15 companies in the Fortune 500 list who had a chief data officer offic but people were trying to you know take that path so that at least the Regulatory and compliance and risk functions can work properly with the right data so that they don't get affected so the auditor or the government regulator should not come back and say your data is wrong so that also led to lot of other consequences for them m so that's
where the things came and uh where it started but then uh this role slowly evolved so in those days it was mostly under the IT team so one person would be you know uh marked as a CDO or some equivalent role in terms of the data management data manager so there uh once this was done and they could clearly demaret that this is the clean data for which we can use for these things then it's SL went that okay we are getting data what else can we do with this and then the business other parts
came in in terms of doing analysis in terms of first you know visualization and generating reports for the senior Executives or for others so that they they get a sense uh in a easier way than just looking at the raw data right so if it's only raw data you see okay last one week two week one month one quarter data it sometimes becomes very overwhelming uh so but people are looking at only key insights to see which is where the dashboards came and then the which led to the evaluation of dashboarding tools also which became
better and better to do that part and then from there came the analytics part of it in terms of doing business analytics or creating data science models so slowly this Chief data officer role has moved more from an IT perspective in terms of just data management to also a business part of it in terms of seeing okay how can we help business uh Executives generate value using the data okay so that's where uh they start showing okay what other value how can we use it you know whether whether it's related to customer analytics or whether
it's related to employee analytics or Supply chainite so all this where what kind of data what cuts of data should I provide to the these users so that you know they can do that so from Pure technical role this is slowly evolving to a more uh business uh focused role and of course with time so some of the companies have also started redefining this roles so calling them like a chief data and analytics officers to nowadays we see quite a few companies having Chief AI officers so who will look at more like AI thing and
then uh take it forward in terms of how to do uh you know those uh AI stuff so that it generates value for the organization so one of the interesting thing you mentioned is that um the rules CDO um has two parts to it one aspect of that is it um corre which is essentially the governance part the security part the organization or the storing part and all that the second aspect of it is driving or deriving intelligence and serve it in a way that um other Business Leaders or lob leaders or maybe the basically
let's just say cxos can make informed or smart decisions right for business growth which means the rules CDO actually need to have both side of um understanding background or maybe even work experience right yes yes so that way it has evolved a lot uh in the sense so now people say they the CDO or whatever equivalent designation you want to give that role should know the data part of things in terms of the data Technologies should know the business so it's not like uh that without knowing the business they can really help the you know
your lines of business of the function heads in that company do anything and third part is now they're saying that person should also know a little bit about AI so that what a use cases can be created out of the data that you have and out of the business needs where do you start what kind of P should you allow this people to do because you can do like thousand different PS with AI but you have limited budget limited time so how do you you know kind of figure out okay this let me start with
this three or four cases do the poc's so unless you know the business unless you know your data unless you know what AI can possibly do you you will not be able to do that correct so that's a good segue to my next question um what kind of challenges cdos face today right considering there are so many variables in the data space right yeah so for cdos again the first and most important challenge Still Remains in terms of uh data quality and data governance and again there there is a uh across organizations uh these things
differ because first and most important question is who owns the data is it the individual functions or the lobs or is it the CDO MH right and there are different opinions and each opinion has its own value in terms of what where they're coming from but again that has to be very clear that whoever is the owner of the data for that part has to ensure that proper data stewardship is happening so that data is clean and data can be used for this things so even if the CD is not the owner of the data
and in many cases probably it makes sense for the CD to not be the owner of the data in large companies uh then she should ensure that you know uh people are at least doing things or you know taking steps to ensure the right correct data clean data is coming to the data warehouse or the lake house whatever from wherever you are picking it up for AI or visualization or analytics so that part is definitely there the second part where it always happens is and this is again is uh in terms of where have you
put this role does it report to the CIO or the CEO or CFO so different organizations again have different things now there is a point of view that it should be under the it under CIO and there's also point of view that it should not be there so I firmly believe that the the St dat officer or equivalent should be away from the it organization primarily because uh it has a very very well defined Charter process and structure right what they have to provide and it's also cost center right so the charter of the CI
is always you have to do this keep the lights on maybe make some changes in the things within this budget or if you can do it even lower but for the CDO he has to show value like what additional value am I getting out of data which is helping the company maybe reduce costs or increase sales or improve customer satisfaction employee satisfaction so for doing that if you may need to spend something right you have to do pocs or you may have to buy some new tools and all so there might always be a conflict
between the two roles correct because one is cost center one is value generating Center correct so that's why it probably makes sense for them to do that secondly for the CD again it's a a thing in terms of uh what I'd like to call is offense versus defense so you you start off from the defensive side in the sense that you ensure that data quality is correct there's no data breach there's no cyber security issues so that is a minimum thing you have to do right it's a non-negotiable hene hygiene and non-negotiable but on the
other side also you have to be on the Forefront and say what things can I do to ensure that you know my business stakeholders are able to see value they're able to do their work in a better faster more efficient way so to uh you know kind of balance these two is also very important uh part of the chief data officer which is why some some companies have now trying to split the roles into two so this data part keeping the data clean and also you know safe from hackers and other things you one person
does chief data and maybe you have a separate Chief AI officer or chief Antics officer to take care of the value creation so again it depends from company to company where they feel you know things may work better and I think it makes sense because majority of the use cases are business focused right so um I think the custodian of data is one aspect of it which incre which you mentioned which means like uh ensuring that um the governance part the storage part the life cycle of that um part which I think is has been
any way a charter on the CIO side of the organization uh the it side of the organization but increasingly the use cases which are more business focused it is important that um the CDO has a skin in the game right and brings in that um Revenue side of um use cases brings in that value um and probably um become a revenue generation function um solely depending on um the management of the data and you know be remain as a cost center corre right and this value generation also is like for this you need a different
set of talent compared to what you have in it it site people may know how to manage let's say snowflake or red Shi you know data warehouse and I have kept it there and it's there and then how do I integrate to different it systems and all those things but on this side you need people who are good business analysts who are good data scientists who can then generate this thing so if I have to do something on the marketing side I have to do personalization I have to do customer segmentation and all those things
so so those are the kind of talent which the regular it team will not have correct so again the role of the c or the chief AI officer becomes to ensure that you get the right talent to do this absolutely so I think um the basically the creation of um responsibilities and the design also sort of sits under that on the application side right right so great I think uh for our first part um you know I guess this was really a Roundup in terms of the the rise of the role of CDO in fact
I was just thinking as you are talking uh explaining the CDO role early 2010s um is when I started hearing maybe between 2012 13 more on um the whole digital transformation initiative right and that actually brought in uh a lot of focus on um eventually CIO side and then you know from there moving on to uh establishing or creating a brand new role CDO uh who played a very important role because the digital transformation again is completely um moving from an offline to an online world right exactly and I think entire 2010 to 20 was
that time or maybe even before 2005 to 2020 I know there are still smaller companies is were kind of making that migration right but we almost take it for granted now that businesses have already moved their um you know many touch points today and online including government right correct most of our government end user functions now looks like our portal based digital app based so on so um just wanted to give a view to um our viewers what's coming down in the pipe in this series what else that we will be talking about just a
quick one or two lines and then maybe we can wrap up this yeah so we will uh probably discuss in terms of how do you structure this whole organization that you building the chief data officer organization the talent and all so whether it's it's it's a single Center Central function and horizontal which cers to all or there are other views that it should be dispersed across all the functions because it's common to all so we will take a look at that what makes sense okay is there a right answer in true sense or is it
also depends on the maturity of the company in the data areas and then you also see if you're doing a central function uh you know how what makes it successful how should it be structured and you know what we have seen some of the more successful companies do in terms of you know examples which has work for them and which could probably work for our viewers also excellent so maybe in this process we'll also talk about a bit of pros and cons of both models right uh and and also maybe take a take an example
or two to narrate this uh in terms of how this is really planning out um in real life as we see in the business right so all rightone thank you very much for your time today uh with this we'll wrap up uh today's episode of the three-part series of building a data organization worldclass data organization and uh we will see you again on the next episode of the same series um for now um we will take a byy from um from you all right so see you again see you bye [Music]
