Latest / The Unpack Show with Anandh / Ep-9: Robin Sutara - AI Capability Leader | Databricks
Transcript
- TheUnpackshow: Hey, everyone. Welcome to another episode. ⁓ In this episode, we are going to cover an interesting topic, I'm passionate about. And I got a special guest who I've met in multiple conferences since he's a good speaker. ⁓ So the topic for today is data strategy in motion. What we want to cover today is what successful AI enabled organizations get right. ⁓ As you see, many organizations talk about data strategy, ⁓ but only few truly live it. ⁓ There's a pragmatic balance between how you do strategy and then how you execute on the strategy. A lot of times I've seen people are very good in strategy, very good in execution, or on both. ⁓ So it is easy to build a roadmap, create a slide deck, or launch a new platform, but real transformation happens when a data strategy moves from planning to practice. from ideas to behavior, from isolated initiatives to a company-wide way of working. So in the era of AI, it is getting more complex. Data is going to be the key. There is no AI without data. That's my personal view. A data strategy cannot just be ⁓ about storing data or managing pipelines. It's about trustworthy data. It's about clear ownership. It's ⁓ culture where experimentation is safe, right? ⁓ the organization that get it right don't treat data as a technical project. They treat it as a leadership discipline, cultural shift, a long-term capability in my view, right? So in this episode, we will unpack those successful, ⁓ what successful AI enabled organization consistently do differently. ⁓ We have a special guest who has gone through this journey, who has gone through that with multiple clients. So I'm more than happy to ⁓ introduce her. But before I turn it over to her, ⁓ The topic is going to be focused on the data strategy in motion, but how the organizations are going to get it right. So Robin, why don't you introduce yourself and then we can get into the topic right away. Robin: Yeah, thank you so much. So lovely to meet you all. Robin Sutara. I am currently ⁓ leading the Americas Enablement Team at Databricks. I've been here for almost four years. I just started this role actually and driving an AI enabled workforce inside of Databricks. So super excited about this topic. Prior to that, I spent the last three and a half years or so as a field CDO here at Databricks where I was working directly with organizations to help them execute against their data strategy and share with them best practices of what was working and what wasn't working. But as you mentioned, right, to the pace of innovation around AI and the capabilities that organizations are looking to deliver and transform their own organizations. I think I was really drawn to the people side and how do we actually change those behaviors? How do we change the culture? Because it's not all about technology, it is more about the people. And so I'm really excited about my recent transition into this role. Prior to Databricks, I spent 25 years at Microsoft where I helped them with their transformation as well, primarily internal around operations efficiencies. TheUnpackshow: Okay. you Robin: ⁓ execution of data strategy inside the company. And so ⁓ I feel like I've come full circle, but always focused on how do we ensure that we're actually driving value? And how do we make sure that we're getting the outcomes that we want? And I think that it goes way beyond technology, which is such a small piece of it. And so super excited to have the conversation today to share with people what I've learned. TheUnpackshow: Mm-hmm. Yeah, thank you. Thank you, Robin. We can get into the first probably discussion right here, right? So we talked a little bit about strategy versus motion. In my experience, lot of companies have a data strategy, but far fewer turn it into day-to-day execution, right? You know, we still talk about legacy data variances, the shadow IT in many aspects, right? But I want to get your perspective, like how you are helping your clients and what's the journey that you went through, right? In your experience, what separates organizations that talk about data strategy from those that actually bring you to life. Robin: Yeah, I think the interesting part is for lots of organizations when you ask about the data strategy, it's one of two things. It's either very, very high level where they just essentially say, we've committed to the board and our shareholders or whoever that might be, that we are going to be data led. they leave it at that big broad statement. And from there, the execution plan. TheUnpackshow: You Robin: isn't always sort of aligned. There is no clarity on what does the change management look like? Who's gonna drive that? What is the operating model, right? Have you actually created out a culture that's going to help you transform? And if I look at the successful transformations I've seen, it is because they focused on how they were actually going to execute from this big broad vision. The other pivot is I often see TheUnpackshow: Yep. Robin: some organizations with their data strategy get so caught up just in the technology. So you look at their data strategy and data strategy is we'll migrate off our legacy data warehouse. it's like, that's not really a strategy. Like why, why if funding got cut and you had to cut a product, right. And things have to reshift the first things that get cut are those things where you can't clearly tie back why that, right. Why that modernization of that data warehouse is so important for the business. TheUnpackshow: Yeah. Robin: What's the value that you're going to drive? If you're a healthcare company, like how is it going to help you deliver better patient experience? If you're right, a school, how is it going to help you educate the students? Like how do you actually tie it back to articulated, clear line of sight between that big vision, the big sort of motion that you're looking to drive, whether it's generating revenue, whether it's saving costs, whether it's driving a culture. right, what, ⁓ helping your employees be more efficient, delivering best or better customer or our citizen experiences, whatever it is, you have to clearly tie the execution of your data strategy to that. And so I think those that have done it really well have been able to bridge those two gaps. didn't just take one ⁓ aspect. They didn't take the big vision only, and they didn't get so caught up into that steps, the Gantt chart of the execution. They figured out how to bridge those things together. So. TheUnpackshow: you Robin: that when they had to pivot resources, when they have to think about ⁓ budget restrictions, whatever it might be that sort of always blows up our strategies, how do they make sure that they have it prioritized in a way that allows them to clearly delineate how they're going to drive value. I think the AI era has also been super interesting. Like you said, without good quality data, there is no AI, right? You're not really going to be able to drive anything. And I think that is becoming, I think, even more critical for organizations to think about their data and AI strategy. TheUnpackshow: you Robin: And how do you draw that clear line of sight into the value proposition, the return on your investment? Because there are lots of organizations that are de-investing now because they didn't do a great job in their strategy of drawing that line between the two of those things. Yeah. TheUnpackshow: Right. Yeah, yeah, since you touched on culture and tools, right? It's an interesting topic. A lot of times we have to put the culture before tools. know, in my experience, every single problem has a solution, which is a technology or a tool, right? We shouldn't be leading with that. It should be an actual solution to the problem, not like in introducing tools or technology to solve a particular problem, right? ⁓ So when you talk about AI readiness, right, we always need to put the culture before tools. Many companies ⁓ try to adopt AI. Robin: Yeah. TheUnpackshow: before building the culture to support it, right? know, bring in tools, bring in vendor products to probably like try it out, which is okay. But you always need to think through the cultural shifts that is going to help, right? Rather than leading to confusion or fear. ⁓ What cultural shifts you have seen help teams ⁓ embrace AI within the organization so that they can win that unlocked large opportunities across the organization? Because when you ask about AI use cases, you can get like... tons and tons of use cases. If you crowdsource it with your employees, you will get locked. But you always need to align it with your priorities. ⁓ And then you always need to make sure that whether it's going to cause confusion or whether it's going to enable, how do you help the teams to embrace the AI rather than just bringing in tools? Robin: Yeah, I think there's a couple of things there. I think for lots of organizations, it is thinking about how do you create a safe space for experimentation? mean, those organizations that I have seen that have been the most successful have realized that there is some level of, ⁓ you know, a build sort of culture, like the business is going to create those things to solve the problem. Like you said, ⁓ use cases is not the limiting factor for most organizations. There are thousands of use cases across the business. The issue is how do you prioritize those and how do you make sure that you're investing in the tools and platforms and capabilities and whatever it is that you need to build to be able to unlock that? And this is why shadow IT often happens, right? Because I think the old legacy world of having a centralized technology team. TheUnpackshow: Yes. Robin: where the request comes from the business into that centralized team and then they have to put it on their backlog and build it out. I think we are seeing more and more of this, whether you call it democratization or self-service or however you articulate it within your organization. Those organizations that are moving the quickest are the ones that have determined the maturity of the business to say, These are the capabilities that you have. have the right, you have the talent and the skills to be able to deliver against these use cases yourself. We will just make sure that you have the right foundations and frameworks in place to be able to go execute against that. So I think this legacy world of a centralized IT team that builds everything and takes requests and takes six to 12 months to build it out for the business. think those days are gone. I think what I see lots of organizations doing is say, okay, how do we make sure that we have the right? TheUnpackshow: Right. Robin: technical and data and AI capabilities and competencies within the business so that we can up level the maturity and the capabilities within that so that they can build against their own use cases. The business leader can make the prioritization decision on where they want to invest, right? What are the use cases they want to invest in those resources? And then how do we make sure that as a platform team or a data team or an AI team, we have the right infrastructure and. governance and all of those things in place to make sure that we're minimizing the risk to the organization, right? And that we're driving ⁓ the moral, ethical standards as well as sort of the quality that our customers and our employees sort of expect us to implement and execute. So I think for most organizations that I have seen that are moving very, very quickly and that they're able to have sort of that ⁓ agility to innovate very, very quickly. They are ⁓ creating sort of this federated sort of model. I know people sometimes call it data fabric or data mesh, and there's lots of contention about what that really means. But for me, what it means is how do you standardize and centralize what you have to for the organization to create the right platforms or foundations in place? TheUnpackshow: Yeah. Robin: And then how do you enable the organization? How do you democratize whether they're a business user, a data scientist or data engineer, whatever they are, how do you make sure that we're giving them the capability to drive the business because they will be able to drive it much, much more quickly. And they have the domain area of expertise, right? That they understand the business value, which again goes back to that strategy, tying strategy to execution. It's going to be the people in the business that are going to be able to articulate that and actually deliver and prioritize against the use cases that help them. deliver against that strategy. TheUnpackshow: Right, so let's assume that the use cases are set, the priorities are set, and we have a clear view of the list of use cases that we want to execute against. But there's a core fundamental problem that I see, and people are realizing it back in 2024 and 2023, like when the AI was hyping up. It's just about data foundation is going to be the key factor. It's going to build the trust and credibility, because without data, you cannot have AI. AI is only as good as the data underneath it, right? On top of it, if you add, you you talked about the shadow IT part of it, there was a lot of ⁓ silos factors that comes into play, right? Business teams, data teams, technology teams work in parallel rather than working together, right? That we see, that's a reality in most of the organizations, right? And so when you put all those into the mix, right, you know, with the AI hype and the data foundation being the critical component to build Robin: Yep. Yep. TheUnpackshow: trust and credibility, having the silos, we need to break the silos so that we can work together, not in parallel, right? ⁓ We need to create a momentum. There's a need for creating a momentum, right, across business, across data teams, across the technology teams. I'm curious to know what is that one ⁓ factor that you have seen working with your clients or in your experience where it really triggered, okay, this small win that we created or the small opportunity that we created. brought in momentum and brought in all teams together so that they can march towards one goal. Robin: I think that I think exactly like you just articulated, I think lots of organizations get lost in the nuanced details as opposed to really ⁓ understanding like what is that North Star? What's the mission? Like, why are we doing it in the first place? Right? Because that's what Galvans does change. If I think about my experience, 25 years at Microsoft, obviously that's long time at one company. But when I started with Microsoft in the late nineties, the mission of the company was a PC on every desktop. When Satya Nadella came in as CEO, that was sort of an irrelevant mission, right? It was no longer applicable. Everybody had PC or they had already moved past PCs to mobile devices or whatever it might've been, right? Like I don't even know my kid knows what a PC is. So I ⁓ sort of think about what he did when he changed the mission of the company to empower every person in organization on the planet to achieve more. Like that was really a galvanizing moment. Everybody at the company could understand like, What was my job? What was my role? mean, at Databricks, our mission is to solve the world's toughest data problems. And so everything I do, every team I work with inside of the company is very much about, does that help us achieve that mission? If it doesn't, then there is no point in chasing that use case. And the minute you can get everybody to see how that use case is going to help them deliver the product, deliver the service. create the experience, work with customers, whatever it might be, how it's helping them solve the world's toughest data problems, that's, think, where you start to get real cultural ⁓ shifts and changes and sort of the pace of innovation that you're looking for. If you leave it, ⁓ if people can't see sort of that driving factor, and people have talked about change management forever, whether you use a... TheUnpackshow: you Right. Robin: know, add car from pro-sci, whether you six sigma, whatever it might be, there's always creating a desire to change, right? Or creating the drive or the knowledge, right? The capability to be able to do it. And I think we've lost sight of that oftentimes because it's all about executing against the change. Can I meet the milestone? Did I get X amount of data moved today? ⁓ The other thing is I think lots of organizations. TheUnpackshow: Right. Robin: ⁓ originally thought about it in a way that it had to be all or nothing. And I think we're now in this era of things are moving so quickly. can't, ⁓ yes, you want trustworthy data, but that doesn't mean that every piece of data that you've had in the company for the last hundred years needs to be rationalized today. Like, how do you think about the real use cases and what subset of data do you really have to focus on to make sure that it is enterprise ready, that it is AI ready, that it is of the quality and caliber that you want to be able to deliver, that you can deliver into the business. Because I think 20 years ago, we were all looking at dashboards and saying, well, I don't have that number. You have that number. Let's argue about the number. We're not rationalized. ⁓ let's just create 700 more dashboards because I don't believe, right, believe your data. I think we are seeing lots of organizations are realizing that that is not a sustainable motion either. And so. TheUnpackshow: Yeah. Robin: really sort of drawing the line in this hand to say, okay, this use case that we've all galvanized around, it's going to help us achieve the organization mission. This is the data that we all agree is what we're going to use for it. right, we all have a commitment in some capacity to make sure that it is of the utmost quality because we're all going to make decisions off of that. And so that's really where we see. TheUnpackshow: Right, right. Robin: ⁓ see organizations that are able to drive sort of that cultural shift across the organization. TheUnpackshow: Yeah, I think you mentioned about this 900 reports and we talked about culture. I don't think we can pass beyond ⁓ this topic when we talk about data, which is governance, right? Sometimes people feel like, ⁓ okay, this blocking factor. In reality, it should be an enabling factor. It should be part of your strategy. It should be part of your execution. But oftentimes people feel like traditional governance feels slow and restrictive, but AI needs speed and agility, right? In a more responsible way. Robin: Hmm. TheUnpackshow: to introduce common language to bring in consistency. ⁓ You might have noticed, I'm sure you might have noticed a lot of organizations standing up AI Council, AI Governance, forums. There's a lot of forums that is popping up right in the organization. But I want to know, how do you lead when you work in a product company, which is enabling the companies who are building the data platform and enabling AI? Robin: Yeah. TheUnpackshow: How do leading organization design governance that supports innovation, right? Without creating a friction or like I would say unnecessary friction. Robin: Yeah, I think ⁓ for those organizations, I think that are doing this really well, what they do is yes, then you have this sort of governing body. I think where most organizations fall down is they don't actually have ⁓ all of the right representation in that governing body, right? Because what will happen typically is, TheUnpackshow: Mm-hmm. Robin: I've seen organizations take just a technology approach or just a legal and compliance approach, right? But how you think about this governing or this governing body is it has to have representation from legal, from HR, from the business, from IT, right? From compliance, you need to have everybody sort of at the table to be able to say, okay, as the governing body, we are not going to make a yes, no decision on every use case. What we are going to do is set up the fundamental principles of what risk factors do we have within our organization? Databricks has a great security framework. Our Phil Cizzo and his team has built out sort of what are the 62 risk factors around AI? And every time I say that for most customers, they worry a little bit. But I always remind them, like, all 62 are not going to apply to your organization, right? This is like every risk. TheUnpackshow: Yes, of course. Robin: factor, it's NIST is using it sort of as their standard. Like it's just a great framework for this governing body where you now have representation across every aspect of the business who should have a say on what are the risk factors that your organization is concerned about. Rather your IREG or your less regulated, like there's lots of factors that are unique to your organization. Look across those 62 factors. The governing body then decides which of those factors are ⁓ specifically relevant for them and which ones now can you put controls and mitigating factors around so that it is no longer a risk factor for you and you can give okay to the business to proceed with pace of innovation they want to, then the governing body only becomes the approver on those youth cases that cannot be right, that you can't pre... ⁓ TheUnpackshow: Yes. Robin: mitigate the risk ahead of time before executing against that use case. And so I think those organizations that have been able to move very, very quickly, they determined what their risk factors were. They built the controls in place to be able to mitigate those risks. And now they're governing body or their council or AI innovation hub or whatever they're calling it, right? Is essentially only taking in the use cases where the body actually has to make a decision otherwise. TheUnpackshow: Mm-hmm. Robin: There's free democratization within the business to proceed forward because you're not having to block every data or AI use case that the organization is trying to drive. Those are the organizations I see that are moving phenomenally quickly. And that's sort of how we do it here at Databricks. Yeah. TheUnpackshow: Right. Right. Yeah, I'm happy to see the shift, right, with respect to what you just talked about. I'm also noticing that data strategy used to be about like access, tools, technology, analytics, reporting. Now it is more about decision-making, providing insights, providing automation, human AI teaming, right? I think it's moving towards that. So as you look towards the next few years or like even the next year or so. Robin: ⁓ Yeah, I always tell people if I knew the answer to that directly so don't make investment decisions based on the commentary because I would be very wealthy if I knew what that looked like but I really do think ⁓ TheUnpackshow: What do you think will define the most successful AI enabled organizations? Robin: We are in sort of this age of transformation, right? I think this is the, and most people have talked about this, right? This is like the next industrial revolution is sort of what we're seeing. What I really, and I think there's lots of trepidation from people on what exactly is it displacing? I mean, what we're really talking about is now I think we have the technical capabilities that we didn't have before where you're really talking about. TheUnpackshow: Yes. Robin: hyper-personalized persona agnostic, integrated into workflow, right, that you can now, technology can now help you deliver against. And so where I'm the most excited is, gosh, I don't know any of us who wouldn't love to get out of this rationalization of 700 Excel spreadsheets, right, to make a business. I really think that is where we will see the biggest impact over this next year is going to be, ⁓ TheUnpackshow: Right. Robin: Automation and simplification of integrated into my workflow that actually helps me drive business faster. So it's not this data overload where you just give me all the data and then I have no idea what I'm looking at. And I'm trying to rationalize it myself, but how do you actually integrate it into what I'm able to deliver? If you think about how long, for example, it would take you to pull together a slide presentation to go deliver content. Now you can put it into an AI generation and you at least have the TheUnpackshow: Yeah. Robin: great work within like 20 minutes and then you can start right and then your creativity can actually drive incremental value and so I'm super excited to see how many sort of repetitive tedious workflows we're now going to be able to mitigate and get off our plate so that human creativity can solve some of these world's toughest data problems. TheUnpackshow: Yeah, I think you're 100 % spot on on the presentation and how the creativity is evolving, right? Rather than doing the repetitive task by ourselves. Like I can understand myself, like, know, preparing the presentation. Yes, you know, it took days and weeks. Now it's like in minutes, right? I think the conversation that we are having is remaining as AI success doesn't start with models, right? It starts with people, starts with processes. Robin: Exactly. TheUnpackshow: And more importantly, the data foundation that makes those models more meaningful. Robin, I think we can talk about this topic forever and ever. But thank you so much for taking the time, Robin, in your busy schedule. And thank you, and I appreciate it. Robin: Yeah, no, thank you. So appreciated.