 Hey everyone, welcome to SuperCloud 5, the battle for AI supremacy. I'm your host, Lisa Martin. Thrilled to welcome back one of our CUBE alumni to the program, Scott Anderson is here. SVP of Product Management and Business Operations at CouchBase. Scott, welcome back to the program, great to see you. Great to be back Lisa, excited to see you again. So CouchBase has some news, I don't want to steal your thunderer. Tell us what's new, what's going on? Yeah, we have a great announcement. One of a series of announcements that we've had over the last few months related to AI. So we've launched what we're calling the Capella Columner service, which is a new service to Capella, which is CouchBase's database as a service offering. It's a new capability that we're incredibly excited about. We've been working with a broad range of our customers in designing this product and really excited to announce the private preview is now available and customers and prospects can go and look at that at our AWS re-invent booth this week. So we're incredibly excited about this new offering. Excellent and congratulations on that by the way. So just back in August, a few months ago, CouchBase announced Capella.IQ, a co-pilot for your database as a service to accelerate developer productivity. Talk about the new Columner service, it will include IQ, give us a double click on that. Yeah, sure. So back at the end of August, we launched Capella IQ, which we're incredibly excited about. And that service allows developers to interact with data and Capella using natural language. It allows them to ask questions, get query definitions, for example, be able to create data, create indexes, all focused on allowing developers to learn CouchBase much more quickly and be incredibly productive when they're interacting with data is part of Capella overall. As we're launching Capella Columner in that new service, we're excited that day one will be supporting IQ. So now in both of our services, Capella overall, which supports our operational database, as well as the Capella Columner service, developers will be able to learn much more rapidly and be much more productive using our co-pilot service, Capella IQ. Give me a little bit of a background. Developers must be just chomping on the bit to this. What's been some of the feedback so far from the developer community in terms of Capella IQ? Yeah, it's interesting. I kind of look at two sets of developers. So developers who may not be familiar with CouchBase that are just becoming familiar with that and their ability to really explore the power of SQL++, which is our query language is part of CouchBase, to understand really, and it's surprising, how similar it is to SQL if you're used to as a relational developer using relational systems. And it's really showed them how they can transfer the skills that they've had for a number of years using relational systems and using SQL and be able to use those skills in the context of CouchBase. The other thing that's really powerful for new developers is some of the operational elements within CouchBase. So how to load data, how to create JSON documents, how to create indexes. And what it's done is really shorten that ramp time to be familiar and confident in using CouchBase. The other set of developers would say are those who are experienced using CouchBase. And they find this incredibly powerful in terms of the ability to write queries and then apply the output of IQ directly by saying, go ahead and execute that code or create sample documents, import those sample documents into a CouchBase database cluster. It also allows those developers to create code in using the standard SDKs that we support, be it .NET or Java and so forth. So not just creating a query in the CouchBase language but being able to write specific code dependent on the language of choice that they're building their application. So it sounds to me like what you're describing, CouchBase has gone right to the developer saying what can we do to make your job easier, more productive and deliver it? Absolutely, and we're really excited about the feedback that we've received so far. In addition, when you get feedback there's great opportunity for us to continue to extend the power of what IQ will be able to deliver. So as we move forward in the coming months and coming quarters, we're investing significantly in this capability as I mentioned before really to increase developer productivity be it somebody new to the CouchBase platform or somebody who's an experienced user within CouchBase. The veterans and the new guides, I love that. So from the research that I've done, Scott, from what I understand the new service will enable customers to do real-time data analysis on the same platform as operational workloads. Talk about why that's a differentiator and a huge benefit. Yeah, thank you for the question, Lisa. It is a huge differentiator. So one of the things that we have done is under the Capella umbrella which is our database as a service we have our operational clusters that we support doing query, indexing, full-text search, KV, a full multi-model database. In addition to that, we now have our new columnar service in Capella. Those are distinct services. One of the things that makes us absolutely unique in the marketplace is our ability to replicate data from our operational data store using memory-to-memory replication. So we're moving data from the operational data store into the columnar service in milliseconds. And we can do that not only from a one-to-one relationship but from a many-to-one relationship. So you can imagine some of our larger customers have several couch-based clusters, tens of couch-based clusters. The ability to replicate the data from each one of those independent clusters into a single columnar service allows them to consolidate their data and get a better view of data that they can perform analytics. The other point that I would call out is by having the two distinct services within the platform is that we don't have any resource contention. So, you know, what you don't want to do is kick off a massive analytical query and impact your operational database that would be impacting the response time of the application. So you're really getting the best of both worlds. Millisecond, response time in terms of replicating that data over, so it's truly real-time, but also the benefit of having workload isolation, which has been a core characteristic of our platform over the last 10 years here at CouchBase. Then we'll look on that from a core differentiation perspective of the CouchBase platform, so the audience understands that. Yeah, so one of the things that we have, and we really call this resource independence, we call it multidimensional scaling. So what that allows a user to be able to do for each one of the services that we have, be it columnar, be it full-text search, be it querying index, be it our data nodes, venting, all the services within our platform. Those services, if a customer wants to, can run co-located on a node if they want, or a specific compute server. They can co-locate only certain services, and as they scale, they can isolate where those services are running on dedicated nodes. And what this allows a customer to do is a couple of things. One is isolate the workload to be able to match the right infrastructure based off of the workload needs within the database. So in some cases, for example, when our data and our index nodes can be memory-intensive. So customers want to use memory-optimized compute for those services. Other things like query are very compute-intensive. So using compute-optimized instances, and being able to mix and match different instance types for those services lowers the overall total cost of ownership, also increases performance, and ensures that we have isolation of those services as a customer scales their overall environment. You touched a little bit on some of the positive business outcomes that I was going to ask you next. Lowered TCO, for example. What are some of the really key benefits that customers are going to be able to glean from the new service? Yeah, let me back up a little bit, and I'll go specifically in the columner. I think some of the benefits they're going to get, and we see this trend overall, Lisa, this concept of highly adaptive applications are real-time adaptive applications. And so the customers that I've spoken with, and we've done a number of customer advisory boards throughout this last year, what they're looking to do is have their applications be more contextualized to the individual user, and be able to customize that application or the content that the application provides to the user based off of real-time information. And not just one or two data parameters, but potentially tens or hundreds of data parameters, so that you and I are getting maybe quite different views from an application perspective as users of that, but it's real-time. It's in the context of me as a user that I want to see. So it's the ability to combine information, not just about my profile, maybe about my order history, maybe geographical information, maybe search history information, and be able to analyze that not just once a week, once a month, and kind of a batch processing, but to do that real-time, minute to minute. So as I'm interacting with the application, the context of the application is changing in real-time. And from a customer perspective, you can look at kind of two sides of the coin. I always look at one which is, how do you drive more revenue as an organization? So that's about presenting the right offers, the right information to me. The other one is driving operational efficiency for the organization itself. So that could be for a field services application, ensuring the employee is getting absolutely the right information that they need at the time that they need it in the context in which they're interacting with the application. That's the goal to deliver better customer outcomes, better customer service, and also reduce the cost of delivering those services. Employee experience, customer experience, to me are always inextricably linked. It sounds like what Capella is delivering is this notion of hyper-personalization, which as consumers, we all expect that whoever we're interacting with is going to be able to deliver relevant content, but it needs to be absolutely specific to me and what I'm looking for. How does that hyper-personalization drive differentiation and value for catch-base? Yeah, let me maybe give you an example, Lisa, one of our customers who's a very, very large retailer in the pizza delivery business around the world. And you can imagine, I think you've experienced this, I've experienced it. We get a promotional offer. The Super Bowl is going to be coming up in a couple of months here. They'll look at my classic order of history of do I buy wings? Do I, what kind of pizza do I like? Do I get the breadsticks and so forth and create a customized offer specific to me at the Super Bowl. That's interesting, we see that today. Now, when I look at hyper-personalization, we look at more real-time context. An example could have been the most recent World Series between the Diamondbacks and the Playoff Game, Diamondbacks and LA Dodgers in which, in the first inning, the Diamondbacks had a pretty quick lead there. And so now I want to present an offer not based off of a known date, right? Super Bowl Sunday is coming up, but based off of real-time information about the score of a game, be able to go ahead and customize an offer based off of my profile being that I'm in, for example, Phoenix, Arizona, or in Scottsdale area that I want to do that. Now, that's interesting. But to give a great customer experience, I need to have enough supply to fulfill that demand. So if I can marry that information with inventory, stocking levels, capacity levels of the individual stores, then I can target the ad, not only to the people of the highest propensity to go ahead and buy, but also my ability to go ahead and fulfill. So if I have, for example, a store that is, now has a significant backlog of orders, I don't want to present a promotional offer to create more of a backlog in my kitchen. So the ability to match that data, you can think of inventory, it could be pizzas, it could also be what's in a store shelves or in an online warehouse that can be delivered, match with an event, a World Series, not even just the World Series or the Playoff, but a specific inning in the game or an activity had with the profile and understanding me as a user and what my preference is, I'm going to get that match of a great offer at the absolute right time with a high confidence in my ability to deliver a great customer experience and fulfill on that promotional offer. That's outstanding. That is, I think an example of real time and hyper personalization at its best. And of course, as consumers, as we are in our personal lives, we expect that these days. It's not a, oh, that would be cool. It's no, that's what I'm looking for. I want to talk now about the acceleration of AI. We've seen so much acceleration the last year alone, but the acceleration of AI does create some challenges for customers. I'd love to understand what those challenges are and how Capella columnar solves those challenges. Yeah, I think there's a number of challenges that we've seen. I think the space is evolving day to day, week to week, month to month. We've seen the changes over the last 12 months in the AI space. With that being said, I think some of the challenges that we hear quite a bit about are the freshness of the data or the real time nature of the data. With Chowchipiti3.5 was released, it was based off of a historical kind of sweep of the internet. What we are seeing now is people want up-to-date information. The ability with our columnar store to be able to feed large language model and integrate with AI DevOps pipelines is critically important. That freshness of the data is absolutely critical. Then really kind of simplifying the overall data processing by having all the data stored in JSON, we think is another very, very important thing. The integrations are going to be absolutely key, but I would say with AI, we're really excited about it from a couch-based perspective because we think it plays into our core strengths, which are scale and performance. So as the amount of data increases, you don't want to have increased latency and response time, and that's one of the things that couch-based kind of a foundational element of our platform, which is hyper-performant, the ability to scale by scaling out the overall environment to be able to deliver that real-time experience to increase the accuracy of the response from things like large language models. Can you kind of double-click on couch-based AI approach its strategy here? Yeah, so there's four key areas that we're focused on in couch-based. The first one we talked about before, which is really around developer productivity. And we think with IQ, we've got a great offering in the marketplace which is going to help developers. And we're really excited about continuing to gain feedback from a broad range of developers and continuing to enhance the capabilities of couch-based IQ or Capella IQ. We're really excited about that. The next thing is optimized AI processing. So this is where speed and performance of our platform comes in. Also enabling AI-driven applications anywhere. And this is a key distinction with couch-based, being able not just to do it the data center in the cloud but at the edge. So one of the unique capabilities that couch-based has is not only we provide a database that can run on-prem in hybrid clouds or in public clouds itself, but our ability to distribute that data with sync capability down to the IoT or the device level. So the ability to connect data from your cloud only sync the relevant information down to your iOS or your iOS mobile device or an IoT device but also the ability as changes of the data occur on that device to efficiently move the data back into the cloud. And as we go forward and talking to a number of analysts and our customers, most data is created on the edge today and that's going to only increase. And we believe AI is going to follow that same paradigm where data is going to be created at the edge and it's going to be consumed at the edge. And that's a critical capability within the couch-based platform. And the last thing that we think is critically important is the ecosystem. So one of the things with couch-based or Capella Colmar that we're introducing is a large number of integration starting with AWS. So the ability to ingest data not only from couch-based operational databases, but from third-party sources such as F3, DynamoDB and other databases into Capella Colmar so that data can be analyzed. You can do calculations on that data. So that ecosystem is critically important and as you've seen Lisa over the last 12 months, that ecosystem has just exploded. The players are going to change. They're going to evolve over time. But we believe it's critically important that we integrate in that ecosystem and we're doing that day one with Capella Colmar. Can you touch a little bit more on the AWS ecosystem and kind of share a bit more on the AWS services that Capella Colmar is leveraging? Yeah, so Capella generally leverages a large number of services from AWS from a control plane perspective, be it secrets manager, obviously the compute, the storage S3. Unique with Colmar, we are using their elastic Kubernetes service. We're using Apache Kafka, the Amazon MSK service also. And then we have a broad number of integrations be it with DynamoDB, DocumentDB, as well as with AWS RDS. So there are a number of services that we're just getting started with and we're really excited about the future ones that we'll be offering as we get closer to general availability. Those will include some of the ML Ops services that we'll be integrating with be it SageMaker or Bedrock, which we think are absolutely critical as part of the service. Wrap us up here, Scott, with some of the main benefits that customers are really going to be able to get from Capella Colmar services. What are some of those key things that they're going to be able to harness right away? Yeah, I will kind of break it down into four kind of areas. The first one is improved agility and performance. And we talked about that workload isolation, the fact that Capella Colmar is a Colmar store, so it's incredibly efficient from a storage consumption standpoint. We've separated compute and storage so they'll be able to dynamically scale that environment out and back in to match the workload needs, which really saves costs. The other thing is the kind of embedded within this stream ingestion. So the ability, as I mentioned before, to bring data in from multiple enterprise data sources and have that data combined and be able to perform real-time analytics within that. Now, one of the most unique capabilities that we haven't touched about yet, Lisa, is the ability not only to do computation within the Colmar store, but take the results of that computation and feed that back into our operational data store. And our data store has world-class performance, the ability to do millions of operations per second in millisecond latency. So the ability to do that analysis in real-time and most importantly, be able to put an operational database, CouchBase and CouchBase Capella, and serve that to the application at low latency and high scale is critically important. The third point I would make, as we talked about many times in this interview so far, is increase ease of use for developer. So IQ once again. And then finally, reducing cost and complexity. The ability to have a single platform that combines an operational data store with the ability to do real-time analytics. It removes cost and friction around ETL, for example, and having multiple vendors and interactions occurring without data. So the ability to have that single platform with those unique capabilities that are isolated, we think is a great benefit to developers and architects in the database space. So many benefits. Scott, thank you so much for joining us on SuperCloud 5, talking about what to do with CouchBase, how you're really enabling hyper-personalization, which we all expect, developer productivity, helping organizations to really leverage the advantages of AI acceleration and reduce things like TCA. We so appreciate your insights. Thank you. Thank you for having me Lisa, really appreciate it. My pleasure. We want to thank you for watching SuperCloud 5, the battle for AI supremacy. I'm Lisa Martin. See you in a few.