Leading from the front: Richard Shaw on How Databricks Thinks About AI Readiness

[Music] Brilliant, welcome to the Databricks Diaries which I showed it's an absolute pleasure to have you on the show. I always like when a Brixta comes on and particularly someone like yourself who's been with the business for, you know, in close to, well, it's just over six, six, six and a half years. You've been with Databricks so, yeah, appreciate your time, appreciate you coming on and sharing your wisdom. I always like to get started by, you know, having the guests sort of do a bit of introduction and sort of whistle stop touring to your background and I guess how you've transitioned into your role and, you know, what today looks like in terms of your roles and responsibilities. Okay, fantastic, thanks Dan, thank you very much for inviting me. I've got some big issues to follow in terms of some of the Brixtas you've already had present. So yes, it's just over six years at Databricks and today I lead the Field Engineering Organization for UK and Ireland and I joined when we were just under a thousand people globally. It's quite large, but now we roughly turn at a half thousand so it's quite some growth both in terms of head count, the size of the business for its performance and also the way it looks as well. That's a really important point to consider. We didn't have a playbook six years ago and this is exactly where we expect it to be. It's a business that is both scaling, but also still has its roots in being a startup in the agility, you know, ambition and how we need to operate. Incredibly exciting place to be. Yeah, absolutely. And if you look back over those six years and go from a thousand employees plus how would you, what sort of experiences has that taught you and what are some of the more challenging times you've had as as a company's grown. One of the reasons I joined Databricks was I wanted to work for a company with a platform that was cloud only. And I've been working for product companies since about 2012 and prior to that on the customer side building and running platforms using the technology of the day. But that reason was I really wanted to try and reduce the amount of time between a customer saying yes, I want to use your product to then seeing the value from it. And so one of the reflections over this six years is Databricks is transformational for many businesses, but that transformation they need to go through can be quite significant. And as much as making the decision on the product and the technology can be can be well managed that transformation they need to go through for their businesses a whole talks to cultural changes talks to skills changes things that often they haven't actually have considered a need a lot of help with. And that angle of hey, you can just pretty much turn it on and go yes, there's more to it than that. But actually behind then transforming their business to really leverage Databricks and what it can do for the round date and I that's another consideration that something we've been learning and certainly in this present age with you know the adoption of AI and where a lot of the products and technology are. That's another consideration on top of a new platform for data analytics. And for people who don't know you rich and you see so you're leading. So you can architecture field engineering. So what does a typical and appreciate they probably isn't a typical day but what what you know what are some of your priorities in your role. To put on with a role in in my business are the solution architect so presales by any other description within field engineering we also have an a brother critical roles like our professional services that sounds outside of my remit. So our solution architecture team are here to help our customers understand who Databricks is as a as a prospective partner to work with because that's how we look at it. It's not vendor customer you know to partnership with them. Don't understand how Databricks is a product could help them solve their business challenges and and achieve their goals. But more broadly we're here to help inspire our customers as well coming back to our partnership piece the way we engage with our customers is not one time activity. You know with their working with them we get to know their businesses intimately and we look on our goal is to build a strategic relationship with them which will be here for many years to come. And that's absolutely true in a large number of our customers where they even predate my time. And we're still working very closely with them today and where they are today is you know not where either a suspect is to be because they're responding to challenges within their business and industry. And they need strong partners to work with to help them to challenge them to work with them. So my team are very much here to to play that role and Databricks is a consumption focused business. So it's not a flurry of activity and then thank you very much we'll see you again in in nine months time that's the traditional sales process in many licensed focused businesses. As is one of continuous engagement. We're helping them evaluate Databricks for one or more use cases and then in parallel the others we're talking about as we look to expand our business and relationship with them. Yeah, yeah it's a very I guess you're quite you're setting quite a unique position because you know you're going to see it from Databricks's perspective. So what kind of common questions and conversations and challenges are you are you discussing you know you know with your customers at the moment. So I'm going to talk about a hot topic and in my role I've become an ex-exponser to a number of accounts. Practically I can't go and visit every single customer but the account teams will ask me if I can come and assist with a few of them. They can be very large strategic accounts as well as smaller developing ones which is really exciting. So I see across a lot of our customers especially around AI is not so much how we understand the technology is a lot of good understanding around that now. It's more what are we going to do with it? How are you meaningfully going to leverage this products to transform or even just apply them to our business. We've got to this this point now where people do understand what agents can do the potential. But there's also the realization that they don't just want to apply agents to a process just because to say hey we've got AI. The opportunity here is to really reconsider how their business operates in an agentic world. Like all these things into a spectrum. Some very advanced customers who have over the last couple of years re-invested time in understanding the technology and working with us as we were with all data producers of products. So they can go through that already evaluation so they can test out product and the technology. One of the first applications was around improving support processes, chatbots etc. As a say space for many companies to evaluate. Now they've got that under belt and they understand what it means to work with AI responsibly think about how they manage the data. They're now looking more broadly at hey where could we actually apply this in our business. But that's not every customer who just expects within an industry you pick 10 customers they're going to be at different levels of maturity. Some practically are just challenged in running their business. Their performance may not be where it needs to be. So they have to think about their core business and less around strategic investments. They still need to do something here because they don't want to find themselves three years down the line getting their business on a sure or a footing but then they're way behind their competitors. Absolutely. And on that so you know, data is got a phenomenal team and you're making strategic acquisitions all the time. So and you're releasing new features, product services. How do you manage that with your customers in terms of that? A amount of choice amount of different products and services available to them. And also the potential sort of lagging time between you know some customers might have just. You know got into data mixed fundamentals or you know, quite new to the product. How do you manage that time and that lag of implementation? You know having your customers benefit from from some of the newer services? I'm engaged with more. to talk about their business and the challenges that they're trying to solve or the goals they're trying to achieve. And we talk about it from a business perspective first and foremost and then we bring in how data bricks can help solve them, which means we're not leading with and these are the new things we can do. They're amazing. It's more actually, let's think about data bricks as a tool to solve that problem. This is how we'd go about it and this is the combination of the products and features we would use. And that really is our engagement model in an up shell, which means it avoids needing to kind of go through those things and see where they might be applicable. We're looking to try and solve the business problem. Yeah. And then that takes us forwards into what combination of products will actually solve that problem and how they evaluate them. And then absolutely, there could be other things they could consider, but it's all around those business use cases because ultimately success for us is the customer sees business value. Not necessarily having all the bells and whistles. It's whatever, whatever is going to bring the most value is the right solution. Yeah, it's not about, hey, you evaluated the entire platform. Great. You've got all the things you can go and you can go and do amazing work with it. That's a build and they will come approach, which does happen sometimes. You'll have a central IT team who want to activate data bricks as an offering within their business. And we see that sometimes successful because they will then have the ability to help migrate from existing legacy platforms to say, right, this is the new one you can use or support you doing that. In other cases, they build the platform and they'll open the doors and expect the business to come running. That doesn't always happen. You then have to go and sell to the business or our irrigation model. And we do that with customers as well. Yeah. Yeah. Okay. And when it comes to the acquisitions that data bricks make, how much of it is around getting there first versus what the demand from the customer? Acquisitions are a strategic move where we may want to accelerate the development of a product or a feature. But it's also around bringing in great talent who could be working on other products and features as well. Hiring is almost important activity. As we continue to grow the business, in all roles we need the very best people to come and support the growth of the business. The product side, sometimes that can mean making an acquisition to bring in a fantastic team besides the IP. Sure. But the number of acquisitions we make is very minor compared to the size of the existing engineering and product team and what they're building. Yeah. It's additive. It's not the main focus, the main approach. Yeah. And having a platform today means we can take these products in a number of different areas. And we're doing that today where it's not only building new products and scaling out horizontally. It's building on top of those products and evolving them. If you look at DLT, where that is now with Lake Flows bar declarative pipelines, that's an evolution. And I think we'll see more of that going forward as well. We'll see this in our partners and competitors. If you look at say AWS, they famously have around 200 products. Some, there's quite a lot of overlap. They're now building some products which sit on top of others, which is very intentional. How many products did they do? Around 15, 16. We'll talk around that number. And it's actually less important to talk around specific products because they have different features. So then you're talking kind of products sets and capabilities. It accurately describes the size of a small tightly integrated and opinionated platform. And that's part of the design ethos from the get-go. Allow us to go and solve any flexibility data that I challenge out there. And I'll try and build end products to say, we've got all these things. But how do we think about a release-trying foundation, which is where we are today with the Unies catalog? And then the different business focused interfaces on top for the different personas. And you touched on talent and hiring the right people and the best people earlier. So how do you internally keep the teams upskilled given that there's a lot of enhancements and improvements to the platform? How can you cover on speed with that? The terrible portman, so internally, is everboarding. Everboarding? Ever-on-boarding. You're forever learning. Which is practically the case. It's a SaaS company. Products are having new features released very, very frequently. Which is really advantageous for our customers. But for our teams, absolutely it's part of their role and the challenge to keep up to date and not just understand what's been updated, what's released, but how they would use it. Comes back to our engagement model. I think it's a very powerful tool. How are we then going to use it to solve this customer challenge? So our team will come from a data engineering with data science backgrounds. So they've all been working on data and our challenges in industry. That's a differentiator for our field compared to other product companies. Any time there's a new feature, they're going to need to know, "Hey, how's that going to help me with this customer? How am I going to build with this?" And that's a major theme for us. As DataBricks is now markedly described as a full-stack platform with the introduction of the DataBricks apps and with agentbricks and with lakebase, we've now only locked another tier of business applications working with our customers. And it's all about building and there will be a ton of unique build opportunities for our team. So it makes a question, what combination of products you can use to solve that business problem and stay on top of it as well. So there's a kind of a meta AI element here. So absolutely, we can read product documents and customer facing product documents. But actually, there's a lot of AI is helping with around the building internally. So we build AI in to every aspect of DataBricks from things like the DataBricks Assistant to actually now saying, "Hey, I want to build an application to do this. What would be the best way?" And why? So now we're talking about a wide coding with our customers. And this is something we're very much doing the team we're doing today. Building rapid prototypes that don't just use say apps and agentbricks and lakebase, but any product to the platform. So in a long-winded way of answering your question, how does the team approach us? They're certainly staying abreast of the product roadmap and the releases. But there's also using AI efficiently to understand what has been built out there and how and what features and then what's the best way. And just, sorry, your character, Richard? I'll just finish off by saying, and that's going to further evolve how our field engineers, our solution architects work with our customers. Not just the pure vibe coding is, but actually looking at how we say abreast of everything in the platform and actually what's the best thing to use when? Okay. And for the layman term and vibe coding, what's the definition? So vibe coding is, in present day, using one of a number of tools to help describe the application or the software that you want to develop and then having a front-tier model usually produce something for you. And even that is moving pretty quickly from iterative prompts to pre-describing some facets of it. So describing it in what are called specs to help guide the model. Massively reducing the amount of time and the level of effort to produce quite significant business applications. Okay. Okay, cool. And then just in terms of the partner network, so I know Databricks have very strong relationships with them every partners. And I believe that number is ever growing. What's the, how do you keep the partners up to today and empower them with the right knowledge to Yeah, you know, against keep up with that evolution and the new releases and the new technologies that are available to them. So it's a one to many relationship we have hundreds thousands of partners out there globally. Regionally that number scales scales down. But then we have different size of partners as well. From the very large GSIs who in region may have thousands or tens of thousands of people working with our customers down to we too size two to ten people, we have ISVs as well. All our partners are incredibly important as part of the ecosystem. So it's a real challenge to engage with them all and do it on a regular basis as our product is evolving quite rapidly. And so we identify Databricks champions and work with our partners to to nominate someone within their business or within a specific team to become that champion. And they will work with us and our partner S.A. team to get together regularly to actually discuss those new features and they are then enabled to then take this to their business. So we have that element scale. So in a large GSI we may have tens of Databricks champions who each play that role. And we support them in sharing that information internally. But the randomly certification, you're going through formal learning and being certified helps validate some of the learning they're understanding. Obviously as the product continues to evolve, the new certifications coming up for new new product. And so here in the London office we have a fantastic event space where we bring in the partner champions on a regular basis to engage with them. Yeah, makes sense. And I think the MVP programs relatively new. I don't know if it was this year or back in the last year that that came out. Could you just tell us a little bit more about the MVP program and what that is incentivising? Yeah, currently we are working with partners to develop, which really run an MVP general program. So focus on some partners. Showing what they can build from a general perspective. Such that we can engage them with our customers to play a role in some of these exciting programs that they're coming about. And much like we're developing our own skills in house. You know, we want to foster some of our partners doing this. And the approach some are taking is coming up with the frame works and accelerators to common business challenges. Or other innovative approaches and applications that they're building. So it's really bringing those types of partners together to see what they're capable of and then understand how and who they're going to who are going to work with us going forwards. Okay, great. And then Richard, if we dig into the sort of AI readiness. So, as you may have seen, we're running on the podcast presently. So I guess from your perspective, when we talk about AI readiness, what, you know, what's your sort of definition and what do you think of on me? Just to say, AI readiness. So coming back to something I talked to earlier. I was expiring a number of our customers and I get to speak to heads of and see level. And themes and the questions. So do I ask you a question, what I think the definition of AI readiness? I break it down in these areas. One is, what are they trying to achieve through adoption of AI? Is it in response to a board directed, so we must have some? Or is it that they've actually identified some programs that they want to evolve with AI? The next one is around skills and the culture. Have they worked enough with today's products to really understand their technology under the hood? And what it can do for them, what it can't do for them as well. There has to be some practical hands-on experience to prove it out. And then the third is around data. There isn't a lot of AI without data. So how are they thinking about their data strategy? Many of our customers are still on that journey. Still transforming a lot of what they do with just the underlying data. But how are they thinking about it for AI specifically? In terms of the data that they want to make available to train or to call on? Have they got that in place so they can move forward with where they want to use AI? Otherwise, that's going to hold them up as well. Now, I'm working with some customers who are putting quite an exciting stake in the ground, which is we want to completely reinvent our business processes using agents. And you know, this is a further evolution from just the application of Gen AI. So there's more learning to go on around how this would work. We believe we have a really strong product offering with agent bricks and the ability to build custom agents, training and data. But it will come back to what are those business processes that you want to focus on? How do you think they could be involved? What do you think the benefits could be as well? That's quite early stage for a lot of our customers. So the business transformation thinking is really, really important. Now, DeadWix is the product company and a partner in many ways. But for a lot of our customers, they need to have a partner who's thinking about that business aspect and that business transformation as well. To make sure they're focusing on the right things and actually they can see success from it as well. The risk that we run is that our customers go, right, we're going to move forwards and do agents. And they spend the next year experimenting and trying. But weren't successful because they didn't have the right business focus on the right areas. They didn't bring the business along with them. If you just transform business process, what impact does that have on the people who can't need work on that process today? How do they work with this? And then if you do this at scale, what does that mean? So it's a very exciting time, but AI readiness, you know, is further evolving itself as quickly as say the frontier models and ideas evolving. The key thing we have customers who want to do this with us, so it comes back to we're partnering with them to help them understand how they move forwards and work with other partners as well who are already advising them to understand how that's actually practically going to work. Interesting. I'm going to shamelessly plug my meetup here because we have our next event in January on the 22nd in Manchester. And two of the three sessions that were running around agent bricks. So we've got one of your specialist solution architect, Dachlan is demoing agent bricks and then we've got a customer who have used agent bricks in one of their solutions. And so they're going to be also demoing and discussing how they've used it. So any customers that you're talking to, which are that are interested in agent bricks. Yeah, please, please send them up north. And exciting show on the 22nd. That's fantastic, Dan. And that very much talks to that that spectrum I referenced earlier on. We've got some who are doing it today and moving forwards and going into production and others who are earlier in the journey. And we need to help them move forwards. And as much for them to remain competitive just to to recognize that are a different phase right now. I give it a, and I don't see the I spoke to last week. Is much more focused on some core business challenges than AI strategy. And he's referring to a conversation where he was like, don't tell me I need one. I've got to sort these things out first. Yes, I am aware of this. This is what my focus is. And as much as we can all just talk about AI, it's very exciting. Some businesses are somewhere on that spectrum. And so we need to kind of think about that accordingly and help them move forwards. And, and, and you know, go through that learning experience, but also understand really how it's going to help them. Because it is transformative today and it will be going forwards, but we're only really scratching the surface in how it's going to transform businesses. Yeah. And just to get your advice on the three sort of topics that you you flagged up. So you've got. I guess the reason, first of all, is like, you know, is it an executive want, you know, is there some sort of fear missing out? We're not doing AI. We need to do AI versus we've got genuine use cases to deliver upon. And then we touched on skills and we touched on the data. So if we go to the start of that, what, how do your conversations differ when you're speaking through. customer who top down, there's the want to have it, but perhaps not the readiness versus there's a clear use case there. How do those conversations differ for yourself? What advice are you giving to your customers on both sides of the coin? A lot of it depends on that point in the journey. Let's say they've got the use case, they've already evaluated Databricks and Agent Bricks and they have their Dejuan place. Then it's a question of scale. How do they now take those learnings that apply that to more use case in their business? We're very confident of where Dejuan Bricks is from a product perspective now and sending Agent Bricks to allow them to leverage agents at scale. Everything would be up on top of them all flow. Everything we have in place with Unity catalog. It's primed. It's their understanding. Let's help them do this, understand which business process is they're going to go after. Where can they now apply agents and do that confidently? The only one is getting them to that point. If they haven't been investing early in our fits now, helping them understand how to evaluate, how to understand really what agents are going to do. There's still a lot of misunderstanding around agents. Being able to work out their way of getting to the output and to the outcome. There are customers out there who think it's just another process. It would just do the thing and just make it faster. There's elements to that, but there's still some fundamentals. There's still some hype and misunderstanding. That comes down to the culture of both the customers. We pay a lot of attention to where is the maturity of the customer, both around not just data, how they're working with Dejuan internally, but also around AI. What's their understanding? What don't they know yet? Are they ready to move to the next stage with us? We have the beauty of working across different customers on that spectrum as well, so we can see what good looks like. We also have to be empathetic to their situation, their culture, where their focus is. Where would you say most people are from? The conversations that you have enriched on that spectrum? Across the entire spectrum, realistically. Like I said, if you pick one industry and pick ten companies in that industry, there will all be at different stages for different reasons. We can only move as fast as our customers do. That's why we have our teams like our delivery solution architects who are there to help the customer move faster, help them with the focus on skills and enable them and help them move forwards confidently. Without those sorts of roles, a lot of it will be up to the customer to try and work it up themselves. They have the products in the hand, but not quite sure what's the right approach. This is really important that they're not just building the skills and confidence, but they're able to demonstrate that internally. It's a lot of concerns around responsible AI and the impact it could have on the business. They're making sure that their own business understands what are the positive aspects and what are the challenges. That's what they have to do to then gain adoption. Typically, we're working with forward-thinking leaders who want to innovate and want to drive the business forwards, but they at some point have to report up to the board around what they are doing and why and how this is working. That's a journey that we're on with them. From what you've seen as well, Richard, why would you say most organisations fall down when they're whilst they're preparing for investment into AI? Data is one. I talked about this earlier around where they are with the maturity around data. Is that like poor data quality, fermented platforms, all of the things? All of the good. Yeah. Because realistically, there are a few companies who've got all of this in place, such as it's now easy to move to the next stage. They have 80% of the data will stay there for a lot or a lot less. They will still have the fragmentation. They may still be on that journey with actually leveraging data bricks now that they've got it available to have all that data in place in terms of the quality and the access. Because to be able to do AI at scale and to really make say, agent bricks available to the business at scale, you've got to have all that ready. You want to make it easier for them to understand which data should it get access to and why. And then people are still working through that. So Unity Catalog, Discoverability, Lineage, Data Quality, Lake Flowsbarb, Declare to Pipelines, all these things, how the customer get there to go. Fantastic. I understand there's a business process here where I think agent bricks could reimagine it. And we can access the data and we're confident around the quality of that data and where it's come from and how it's getting to us as well. Great we can move forward. So if you think about a business with tens of thousands of users where they'd like to use at scale, there could be end data sets that they need access to. And it's so much easier if all of that is in place. But everything happens in stages. So we have customers using agent bricks today. And in a year's time, I'm anticipating and we're going to have customers who are building quite complex networks of agents to solve business processes that call on multiple data sets. These agents are interacting with MCP with agents on other platforms as well. So looking forwards, it's going to be a network of agents supported by data and network of agents that's across platforms as well. So that's another level. And practically it's a tool today. But not many businesses are doing at scale. And that's where it really counts. You absolutely get the forward thinkers who innovate, but they'll be in a small minority. The vast majority of businesses are back here and our journey is to help everybody move forwards. Yeah. You see that a lot of organisations currently are using AI for administrative functionality rather than truly value creation. So how could an organisation or what and vice would you give to an organisation to try and change that from using it as a helping hand to review or type of notes or summarize a call to actually change in the culture of a business and to become a revenue generator for them? Well, that's where my team come in. That's where it's my team and others to come and discuss what some of those business processes are, how the business is working, what some of the outcomes you try and achieve and work back from that as well as look at new and innovative approaches where agents and Jenny I could support. That's the really fun part of my team is rather doing that discovery and talking with the customer because in most situations the besides the kind of low hanging fruit you've described there is a lot of others where customers are not sure or they have an idea but they'd really benefit from hearing from my team around what we've done with other customers in this industry or other industries approaches that could be applicable. That's the stuff that really gets me excited where we bought a use case in that they'd never thought about before as another approach to achieving something. And with agents and Jenny I this is still so much discovery and experimentation to do so that's the partnership that we really bring besides the world class day-to-day platform. It's engaging with our customers and helping them think differently because we have the benefit of coming in and not being part of their business but knowing a lot about it and having a different point of view and saying "oh but if you thought about doing it this way and you could leverage Databricks to do this that he has a lot of before that's how we help the customer move forward." Yeah very good and for business leaders who are listening to this podcast who perhaps don't feel they're quite ready for some of the more advanced technologies and features that are available to them today. What would you advise them other sort of priority things for them to you know maybe the top three priorities that they should take. I had a further investment. So before further investment in AI, it's going to be the data, the data, the data, there's your three things in one. Because it will only support that investment, that experimentation, that transformation as well. It will give them confidence around what they're doing. It will help them leverage these new products and features for their business. They don't have that in place. They will struggle. They will move forwards and want to do more, but won't have everything in place. And then they'll be held up whilst they get the data in place. So it may sound a bit like a boring answer, but it's a realistic one. And that was three in one, but really the two investments in some experimentation. You know, to build the confidence and the understanding to then move forwards. I've seen it time and time again with our customers who have done this early on and have gone past some of the low hanging fruit, to build the confidence, to get the support from the business, to go to that next level. You can't jump straight from one to the other. You've got to go on that journey, so make sure you are investing. So for a lot of our customers, like I gave an example of who are focusing on more normal business challenges, you may not have the AI strategy. They do have people near to investing and experimenting to get that experience to, like you can move forwards, otherwise you'll find yourself at a disadvantage. How big of a challenge is the backing of AI within your customers at the moment, in terms of people willing to experiment, you know, and take a bit of a punt, I suppose? There are a lot of people who want to, but they may not have the remit in their role to do that. You know, they might be limited by the access that they have. You know, if a business hasn't signed off the use of AI and is comfortable, they may not be able to, especially within regulated industries. So it's not something that any company could ignore, but they need to make sure they've got groups of people, investing time, developing the skills. So their minimum is working with partners to actually show what's possible, and engage in a regular basis to explore and build that confidence and understanding. So this is having partners come and do, like, prove a concept, pilots. Yeah. Exactly. Yeah. Interesting. Yeah, brilliant. I mean, we've covered a whole host of topics across the show. I think it's been really fascinating, as I say, I was excited to have you on, because, you know, you're seeing it from multiple different lenses. I suppose my last question would be how, how would, how are data bricks themselves internally dealing with the rapid enhancements and the demand and the hunger for AI? What's your own internal take? We're at a, we're at a point now where there's going to be quite a sizeable, not change, but evolution at what we're doing, certainly within the field engineering or going forwards. And what's driving that is both our own products, the market maturity around AI, and the need to leverage more of the technology products and our project products are out there to help us do what we're doing. So even over the last few months, we've been running a global program to identify and develop new IP internally. New tools and systems, which can help us. And it follows the theme of building. So I talked earlier around vibe coding, and that's playing a really important part in how we build internal tools to work with our own data. But also gets the team excited around what they could be building and building a lot faster. Within data bricks, like I said, AI is prevalent in the use of the product from the natural language, kind of the system, to describing what we want to understand of the data and how to present that. That's there for all of our customers to use. And we dog food our own product, we use it every single day. But it's now, let's explore what we could build that we haven't got today that would really help us. So within the field and community, there's some great projects spring up, where we're developing custom applications that can help us do what we do, or can help customers, or even some custom applications that we think, "Hey, this is how we sold the customer product previously. This is how we do it in the present day, especially unlocking data bricks apps and agents and late-base." So how we're using AI today is quite similar to be different from where it was a year ago. And where you're just getting started, this is really really exciting, is it again talks to your other question, "How do we keep the team abreast of what's being released?" Because they're building with it on a daily basis. The benefit of our business and for the benefit of our customers as well. And our local data bricks now, like I said, is a four-step platform, practically, obviously there are some limitations. Practically, there are no data and AI challenges that we couldn't solve with the platform today. Yeah. So a lot of internal experimentation going on. Correct. Yeah. Fantastic. Well, thank you, Richard. I really appreciate your insights and your time. I hope you enjoyed the discussion. But yeah, great to have you on the show. And yeah, excited to see what you guys announced next year. I always, well, I have done for the last three years now, been over to San Francisco for the summit. And it's always exciting to see what you guys are releasing. So I will be paying close attention. Thanks, Dan. Really enjoyed the conversation and for being invited to give my perspective on our world. And I think we'll probably stick it around 15 products as the official without starting to count them too much. I'll close with this. Databricks is a fantastic platform, but it's really a tool there to help solve our customers. Challenges and help achieve their goals. And I'm a technologist. I've been working on technology in my career. It's the application of the technology that gets me excited. What did we help and do with it? What did we help achieve? And look, you'll have seen we continue to innovate both through what we develop. What we bring to market and also what we and who we acquire as well. So we firmly believe that if we continue innovating, bringing a really strong product to market, it will continue to drive our growth. Our customers will really enjoy using Databricks to help them. Absolutely. Particularly if it's having value. Absolutely. Brilliant. We'll have a lovely Christmas. Yeah, catch you soon. Thanks a lot. Thanks Richard. Cheers. Bye-bye.