Vanguard Chief Data Analytics Officer Ryan Swann on Boosting Analytics Performance

foreign speaks we're talking about a method to improve your data analytics practice by using a centralized co-located data and analytics team to discuss that I'm joined by a major industry expert with me is Ryan Swann Chief data analytics officer at Vanguard Ryan thanks for joining us today thanks James thanks for having me looking forward to the chat so I want to talk about the the co-located team idea that's interesting but before we even get to that I think it's kind of a larger sense of what's going on in the analytics profession these days as you speak with other data data analytics professionals what are a couple of key trends and or challenges you see with how companies are are using data analytics these days seems like that the dream doesn't always come true when it comes to the analytics practice uh yeah that's a great question James listen across the industry more and more firms are trying to use data and analytics Ai and ml to help draw business outcomes reduce risk and that means that a strong data Foundation has to be established so you need a mature ecosystem you need a data culture where uh where employees and contractors and Consultants really understand that data is a strategic asset it has to be governed in and managed with a high level of quality so that when we use it to create additional data build or train AI models uh that your your result is a high quality result and so I really across the industry I think my peers and myself are always trying to move our organizations forward to to to establish that foundation and scale it while we unlock the value of analytics um AI in the mail uh at the same time and so it is uh you know it's a it's a constant Improvement uh and a job is is never really done because we continue to create more data we continue to um identify additional resources additional correlations uh that that bring us Insight um about how to serve our clients so uh and so that's what I'm saying today yeah it's interesting well you've spoken about the importance of a centralized co-located data and analytics team first let's define this what does it mean to be both centralized and co-located yeah so so typically uh organization data and analytics uh there's a pendulum if you will on one end it's the siled fully embedded kind of uh fully embedded side-loaded nature if you will data and analytics scenes versus a a totally centralized data analytics team that is that is providing uh products and services to the business lines and typically organizations particularly Enterprise organizations are somewhere in between uh and so what we mean by essentialize but co-located is that when we set up the cdao as we were trying to continue to scale the use of data and analytics here at Vanguard there were there's some there was some things that we wanted to keep with um a kind of Federated or a hub and smoke model which is context matters now we don't do data in analytics or Ai and ml for data and analytics sake we do it to take a stand for all investors and give them the best chance for investment success so so the content the business content matters and so when we say co-located what we mean is that we want our data analysts uh and our data professionals our data Sciences to be as close to the business as possible we want to align their objective so to the business priorities are their priorities right and so that they get the context of while we're creating a dashboard while we're creating an AI or ml model a propensity model or natural language processing model but it's centralized in the sense that we're able to help give that those crew members that's what we call our employees to give those crew members um career Pathways we're able to help them develop in ways by looking horizontally and and and partnering senior data scientists with Juniors data scientists that may be that may be supporting different business lives and so decentralization allows you to to provide a training and development career pathing uh environment while you enable Enterprise capabilities and kind of cross-functional sharing of best practices that you that you don't necessarily get when um when the pendulum is on the far on the far side of um The Far Side of of being like being siled and that type of nature so so that's what we mean when we say centralized but co-located it really allows us to get the best of our worlds it seems like and correct me if I'm wrong the co-located piece of that means that more people get to access the data because it could be in many different locations but the centralized nature of that is better for governance am I correctly thinking that um yeah partly the the co-location uh literally means that the data sciences that the data analyst is like sitting in the same pod it's like on the same floor that the person is sitting next to them is probably someone in in the business versus in the cdao um we do from a a data access perspective or data Lake perspective we do have Enterprise uh storage and compute that allows us to leverage a scale inefficiencies to access data uh uh compute uh storage those types of things we we absolutely do that as well but the the co-location is is more meant to be hey I want them to be a part of the team right I want them to be a part of that business team that is driving business outcomes whether it's for our institutional investors our retail investors or providing advice uh for clients Financial advice for our clients you know you've talked about this somewhat I like to sort of drill into this topic I mean how does this model improve functionality how does it does it help the company or any any company that uses it yeah so I think that a couple of things from a people process technology standpoint it helps affirm build data accliment across a number of organization a number of professions right if if this if the data analytics uh professionals are all centralized and kind of often their own building not not actually their day-to-day with with the the business Specialists or the policy analyst or um the sales executive then the Net Connection is is harder to make not impossible but harder to make and so and so that's a key benefit because you're able to connect the dots stratum offense if you will between the technical side of what we do and the why it matters of what we do and that allows uh business Executives business managers Frontline managers sales Executives to understand how they are operating impact how they are operating impacts data quality data as an asset and ultimately our predictive power to help draw business outcomes right and so by connecting those thoughts you kind of are raising all votes and then from a technology perspective you're able to take capabilities that take capabilities that may have been developed in one business unit and use them in another business unit or find synergies and correlations horizontally across the firm that'll allow you to unlock value that that you didn't anticipate and so we've created a significant amount of value incremental value um across the firm by doing just that and as we continue to to to to mature and kind of move up that the data maturity curve um it just becomes more and more important that connective tissue between seeing the data and analytics and in the business why it matters and so the co-location is a vital part of that well I think the other important piece is the idea of developing talent and then sometimes those Junior analysts could use the help of mentorship I mean does this help develop Talent among the data analytics stuff absolutely absolutely you gotta you gotta put yourself in the shoes of the front line the front line employee or crew member so so they need they need to see a career path for themselves if if you if they look around and there's only one or two data analysts in a business unit then then they know for career development perspective it signals that hey in order to maybe get to a manager to get to a senior manager you may have to change professions and that's not what we want right what we want is to create a way to give them a career path and show them opportunities of of growth maybe you want to go become a technical me or maybe you want to become more of a general manager and so the cdao gives them that gives them that those Pathways and that comes with like you mentioned mentorship that comes with training and development learning new skills maybe you know maybe you know SQL and you want to Learn Python um maybe maybe you're you're a data analyst and you want to become a data scientist and so those Pathways allow us to to continue to develop and and and encourage our crew to learn and grow um so that that as they grow as they move through the organization as they grow in their career they have multiple opportunities uh to uh either go technically or to grow technically or to to go more into leadership and so it's been very effective our engagement of our data analysis crew has significantly increased uh and I think that's uh that's that's a that's a big reason right every company uh in the industry uh particularly in financial services and National management is trying to do more with data and analytics so it's very important to us that we help our crew grow we we we we challenged and we're challenging Pro uh uh uh projects that are actually making a difference uh in the lives of investors and helping them achieve investor the investment success and in that career development is one of the ways that we do it um that matters for sure uh I think the big question a lot of companies are wondering is you know where is this all going in the years ahead I mean do you have a sense of future directions for data analytics practice in a corporate setting what do you foresee absolutely I a number of things are happening I think from a AI nml perspective I think you will continue to see more and more AI capabilities generative AI capabilities leveraged by firms across the industry I think you'll also see uh and you're seeing this already where SAS offerings of technology that the Enterprise technology incorporate gen AI or AI capabilities directly into directly into their product whether that be something as simple as Microsoft Office or our Salesforce or Tableau any of uh uh I could just go on a list of companies that are all have signaled that hey we want to deploy gen AI capabilities into our into our software that's going to help us be more productive it's going to help us identify insights that we that we that we would not have seen before and so that means from a data and analytics perspective not only do we have to continue to help unlock that value you but there's a offense and defense to this right which is there are significant risks with leveraging generative Ai and AI capabilities that we have to govern for um and so in the cdao at Vanguard our strategy is both offense and defense where we are bringing together um both a capability that's going to unlock value but also reduce risk in a proactive way and so what I mean by that is things like privacy by Design model governance things like um Improvement accuracy to reduce hallucination uh uh in in some of these models and those capabilities ultimately allow us to build functionality that will at scale um have the right risk appetite if you will um and so that's that's very important and I'm seeing that across multiple Industries particularly the regulated industry so so Finance Health Care Pharma defense those type of Industries where uh where uh there's a lot of scrutiny right I think you said it it's going to be an interesting sector to follow uh thank you so much for sharing your expertise today and please come back and talk to us again sometime awesome thanks James thanks for having me

This transcript was generated automatically from the video's captions and may contain errors.