MongoDB, Inc. (MDB) Earnings Call Transcript & Summary

Please welcome to the stage MongoDB, President and CEO, Dev Ittycheria.

So you guys thought you're coming to a developer conference, Little did you know you're going to New York City's first rave party for today. So good morning, and welcome to MongoDB 2022. It's great to be here. It's great to be back in New York City. I cannot think of a better city to reconnect. I love New York, the energy, the vibrancy, it really makes New York a very special place. We're super excited to be here with all of you today. I know on behalf of the whole company, this is something we're looking forward to for the whole year, and we have a great show in front of us. The last time we actually got together was in 2019, 3 years ago. And I know we all know why. And we all have our own stories about COVID, and we surely have ours. I remember that faithful Zoom call in the middle of March 2020 when I got together with the senior executive team in the company, and we're trying to figure out what the hell to do. The world was shutting down. There seems to be a lot of panic. We weren't -- no one had seen anything like this ever before. And frankly, we wondered if anyone even would care about what we're doing. So the implications were not just disturbing, but in some ways, almost terrifying. But the funny thing happened, it turned out that people did care. Because in a world that was locked down every organization, every business had to reorient itself to connect with its customers and partners and suppliers and other stakeholders through digital means. What it meant that everyone had to have a digital-first strategy. Every management team, every board, every company was asking themselves, how quickly can we pivot to a digital-first strategy. Retailers have to become e-commerce experts overnight. Gaming and streaming companies had to suddenly deal with this influx of demand they had never seen before. Pharmacies have to figure out how they would get medications to their customers when there was no 1 in their stores. It was a pretty dynamic time. Time was critical. Speed was essential. And what people quickly learned was that the fastest way to build was to use MongoDB. The fastest way to build digital capabilities was to use MongoDB. Developers can move insanely fast and ship quickly. Product leaders to get their products out the door faster than the competition and C levels could quickly figure out how to seize new opportunities or respond to new threats. So why did MongoDB make it so easy to go fast, 2 words, developer productivity. This company was founded day 1 on enabling developers to be highly productive, and it remains that our ethos starting day 1 remains consistent today and remain consistent tomorrow. Every product we build, every feature we develop is all geared towards developer productivity. So the obvious question is, how do you make developers insanely fast and productive? Well, you have to ask, where do developers spend most of the time? They spend most of their time working with data. So if you remove the friction with working with data, you make developers insanely productive, and we have shown that we remove enormous friction from the workload that developers have. And when you remove that friction, you know what else you do, you help organizations innovate more quickly. Now I've been C of this company for almost 8 years, and I've met with thousands of customers all around the world North America, Europe, Asia, everywhere, almost everywhere. And the customers have told me lots of things. They share with me their insights, their challenges, their issues, they're going to be feedback when we haven't done exactly what they wanted. Well, one thing they've never told me, not 1 single customer told me this, no customers complained about innovating too quickly. What they have complained about and what they struggle with, how do they increase their pace of innovation. And invariably, the thing that's hold them back is their legacy, brittle and inflexible architecture and infrastructure. So that's exactly why MongoDB became so popular. And MongoDB was built on the notion of a document model, where you could completely think -- transform the way you organize and manage data. Now there were many naysayers out there across the industry who challenge the notion that the document model could truly be a mission-critical data model to enable the wide variety of use cases. There might be even some today. Well, developers spoke and said something different. Our software has been downloaded 265 million times. To put that number in perspective, our software has been downloaded roughly 10x for every developer out in the world. And you know what's even more interesting? This year alone, our software download more times than in the first 12 years of the company's history. MongoDB has never been so popular. Now if you've been following the company, you've heard about a product called Atlas. Atlas has truly been an inflection point for our business. The majority of our customers use Atlas. Most of our revenue comes from Atlas and Atlas is the fastest-growing piece of our business. You're going to hear a lot about Atlas and all the innovations that we're introducing today and over the next 3 days. And there's going to be tons of really interesting breakout sessions that we'll go into a lot more technical detail. But just to give you a little bit of perspective of how far we come. In 2016, we introduced Atlas on 1 cloud on 4 regions. Now a lot of people thought we were crazy to try and partner and compete with the cloud providers. No company at that point in time had ever built a business of any substantial scale trying to do so. And we got a lot of naysayers who said, "This is crazy." By 2017, we launched a free tier, and we're adding about 2,500 users per week. Well, fast forward today, Atlas has now spans nearly 100 regions around the world, across 3 different cloud providers. Not only does Atlas offer a great range of features, it's also a true multi-cloud service. We recently opened regions in Indonesia, Chile, Sweden and South Africa, and we're going to keep doing so. And oh, by the way, that free tier, we're now adding 25,000 users per week. So by any measure, Atlas is the most widely available cloud data service on the planet. Not only do we offer a great service, we actually have great relationships with the cloud providers Amazon, Azure, Google are here at the show today and over the next 3 days. And we've built deep integrations with their products and also partner with them on both the marketing and sales side to help customers drive value. So what does this mean about customers? What are the implications? Our first show was -- our first world conference was in 2014 here in New York City at a lot smaller place. At that time, we had about 1,100 customers. Today, we have 35,000 customers, some of the most demanding and sophisticated customers around the world across nearly every industry every geography, every customer segment using us for almost every conceivable use case. Now you're going to hear a lot of our customers over the next 3 days, and we're thrilled to have so many customers who are going to speak to you about their journey with MongoDB. But let me just highlight just a few. Boots was a health and beauty retailer in the United Kingdom. They're a 180-year-old company, 85% of the U.K. population is within 10 minutes of a Boots location. When the COVID hit, they had no way to get medications to their customers who desperately needed them. Guess what? They pivoted really quickly, built a digital platform on MongoDB and essentially help the customers to get good health care. Toyota, the whole connected car program is built on MongoDB. So all the telematics and sensor data that's collected from the cars is collected through Atlas and sent to the service centers and as well as to end customers to track how my car is doing as well as to predict when the car needs maintenance. A company called QHealth completely transformed COVID testing. I'm sure all of us have gone to some clinic with lots of sick people, hoping that we don't get sick, while we want to get tested. Well, QHealth basically enabled PCR-like testing in the privacy and safety of your own home in less than 20 minutes, all -- and they built their platform on MongoDB. Verizon is basically using MongoDB to move data closer to the edge to drive the next generation of low latency applications that will be delivered over 5G. We have so many customers doing so many interesting things. It's amazing. In fact, I would warrant that all of you have already interacted with some customer using MongoDB before you got here, and you'll do multiple times the rest of the day. There's 1 customer I'd like to introduce at this stage is Wells Fargo, a very well-known consumer bank. I'm really pleased to introduce Catherine Li, Senior Vice President and digital name to the stage. Please join me in welcoming her.

Thank you for your warm introduction, Dev. Good morning, everyone. My name is Catherine Li. I lead digital enablement for Wells Fargo Home Lending Technology. Today, I'm going to talk to you a little bit about change. This year marks 170 years of funding of Wells Fargo, Henry Wells and William Fargo created this innovative little startup to help customers do 2 simple things: manage their money and create new businesses. As you probably can imagine that customer has changed immensely and that the needs of that customer have changed immensely as well. We must adopt and adapt new and modern technologies to serve the customer and to service that changing customer needs better. So changes force us into transformation and the pandemic has changed, customers significantly. According to Accenture Research that today, there are more than ever reimagined customers -- reimagine the customers more than ever before. Those customers are willing to pay more for better services and products and for brands they care about, and the brands are more aligned to their whole values. The transformation in customers' behavior is forcing companies change or to be changed. At Wells Fargo, we've had asked ourselves, are we delivering the right products through this reimagined customers? Our changes must be always informed by what customers are looking for. By looking at our user data of those remain customers, we saw that customers wanted to be empowered and to better understand their credit score and to be able to take actions confidently about their credit score. We have chosen to change by investing and partnering with MongoDB to get those insights more than ever before and faster. Today, the amount of data that companies store to understand what is changing about their customer exponentially larger than before. And then last year, this time, we rely on MongoDB for its speed in development and the responsiveness in change to get those insights and to change ourselves even faster. As a response to Wells Fargo customers, we introduced Credit Close-Up, a digital online product that to enable those customers' needs. Since its launch, millions of online customers have opted in for us to continue to scale products and services for customer base that we have, we need MongoDB, building resiliency and scalability to continue to be successful. Our customer base is not just very large. It's very diverse. Last change or final change I'm going to talk about today perhaps, I should say, first change, we have been seeing and has been happening is our workforce. As a female Chinese engineer, I'm very, very proud to be 1 of many faces in today's diverse technology team at Wells Fargo. Our team reflects diverse and ever-changing customers we serve and it's really critical that we design -- and design and develop solutions to serve those customers. Today, customers are not just right. They are taking control of their wallets and ultimately decide who they spend money with and a few questions I would like to you all consider today. Do you fully understand your customers changing? Do you fully understand the change taking place with your customer base? And are you confident that you're responding to that change in the right way, managing, measuring and generating insights from your data to answer these questions, these type of questions is critically important. I want to leave you with a famous saying by Peter Drucker, "if you cannot measure it, you cannot improve it." Thank you. I'm so honored to be here today.

Thank you, Catherine. The Wells Fargo story really reinforces the notion that to build a durable business, you need to adapt to change quickly. And just to put things in perspective, Wells Fargo started as a small MongoDB customer a few years ago, and now there are hundreds of MongoDB applications running all aspects of the bank. And candidly, we're so grateful to customers like Wells Fargo and the other 35,000 customers who've enabled our success and our growth. We would not be here today without them. And our success and growth has enabled us to invest back into the business and make our products even better. In our first conference in 2014, we had invested up to that point about $73 million in R&D. Today, that number is close to $1 billion. So we feel really proud about how far we've come in a short period of time. But folks, we're just getting started. Software is transforming almost every facet of our personal and professional lives. IDC says that there will be 750 million new applications by 2025. What's even more interesting there'll be more applications built over the next few years that were built in the first 40 years of software's history. To address this increasing demand, we are taking this very, very seriously. And consequently, we're investing more in R&D this year than in the first 10 years of the company's history. And the reason we're doing so is to our mission statement, our mission is to really empower innovators, integrators like people in this room, innovators all around the world to create, transform and disrupt industries by unleashing the power of software and data. And we have a lot of confidence that we can make an impact because our world view is informed by experience over the last 15 years. We know that relational databases have way too many limitations. They're too rigid, too flexible, too cumbersome, too costly and just don't scale. And other people know that, too. So there's been a proliferation of niche databases who are basically focused on some small point solution to compensate for the limitations of the relational database. But the problem with using more and more niche databases, it creates more and more cost and complexity to your architecture and you get an architecture that looks like this, multiple technologies, a very disjointed appointed developer experience, data silos all over your enterprise and high cost to learn, support and manage this infrastructure. Who the hell wants an architecture like this. And you know what it also slows down innovation. So the market needs something different. We believe that developers need broad support for a wide range of use cases from operational to transactional to analytical use cases. We believe that development teams need to be able to service more of the data life cycle to be able to analyze to transform and move data without using batch processes or ETL jobs. We believe that developers need to -- want to build on a modern data model that's designed to the way they think and the way they code. And we also believe that developers want an elegant developer experience that makes their lives so much easier, and they want all this in 1 unified platform. What they need is a developer data platform. No one else is delivering a single unified developer-oriented data platform in the industry like Atlas. You're going to hear so much more about this over the next 3 days. Our next speaker, our Chief Product Officer, Sahir Azam, will go into a lot more detail as well as talk about all the new product innovations that are coming out. And later this afternoon, our CTO, Mark Porter will also share some new product announcements as well as talk about why MongoDB is so profoundly different than any other technology in the industry. I'm really excited about the show that we have for all of you. We have a jam pack 3 days. There's going to be over 100 breakout sessions, again, going to deep technical detail because I know there's a lot of people who want to really get into the technical guts of what we're doing. There's going to be over 50 customer presentations. We're going to have futurist, Ray Kurzweil, talk about whether industry and frankly, where the world is going. And oh, by the way, we can't actually have that rave party tomorrow night because we have a little fun tomorrow night. So please join us. Lastly, I want to thank our sponsors that have been really important to helping our growth and play a big role in expanding the ecosystem that we work in. And most importantly, thank you to our premium gold sponsors, AWS, Confluent, Google Cloud and IBM. So have a great show, and please join me in welcoming Sahir Azam, our Chief Product Officer to the stage. Thank you.

Good morning, New York City. We've got a packed house here. And I want to give a shout out to all the people listening remotely for all the new announcements coming today. As Dave mentioned, our mission at MongoDB is to help innovators create, transform and disrupt industries by unleashing the power of software and data. And when we refer to innovators, we're talking about all of you, whether you're helping transform your existing organizations or building a new -- create new product, you're changing the way the world works and building the experiences of tomorrow. And we're here to help you do it. So as we think about the product strategy at MongoDB, we obsess about our focus on helping you get from an idea to a global reality, end to end. And we do this in 3 key ways. We start first with a focus on an elegant developer experience. We want data and the plumbing and the complexity to get out of the way so you can focus on innovating and building the differentiation for your companies and your ideas. We enable broad workload support, so we can support most, if not all of the capabilities you need for your demanding modern applications end to end. And we do this all with a strong foundation of resiliency, security and scale. So you can take an idea from a single geography to serving customers worldwide. And that's why we built Atlas, the industry's first developer data platform. It's built around the document model, unleashing innovation and agility in the way you work with data. It gives you the broad capabilities from search, operational, transactional and analytical capabilities end-to-end in a single platform, and it has leading capabilities around multi-cloud resilience and security with all the brand new announcements we'll talk about today. So let's get started. We're happy to talk about all the different things we've been working on for the last year. So starting first with the elegant developer experience. Modern development in many ways has been a constant surge for higher levels of abstraction. And that abstraction has been built to get complexity out of the way, so you can move faster, differentiate and pivot if needed. And this has been happening all across the technology stack. We've seen compute move from physical machines to virtual machines, and that's given way to containers and APIs and serverless platforms living up to the dream of the cloud, allowing everything to be a true utility. And in fact, this way of working is only expected to grow. Analysts project that this is going to be an almost $40 billion market in the coming years. But as we surveyed the data landscape and thought about how serverless databases should be built, we looked at the existing models in the industry. And we saw that many of them imposed some significant limitations. For one, the most popular serverless databases aren't really full-feature databases at all. They're often key value stores, which forced you to build on more technology, more complexity to serve the needs of your modern applications. Or they're shackled with a typical rigid traditional relational model that holds you back from moving fast in the first place and compromises about the elastic scale that the cloud actually -- the cloud actually promises. So last year, we were super excited to release the preview of Atlas serverless, which brings all the characteristics of a serverless fundamentals forward in MongoDB had experience. One, minimal configuration, so you can get up and running in seconds; two, seamless scaling, whether you're scaling up rapidly because you're opts are super successful or you're scaling back down because your workloads more batch oriented or sparse overnight. And of course, usage-based pricing. So you only pay for what you use. Of course, Atlas serverless brings the full power of the MongoDB experience. The rich document model, transactional guarantees, rich aggregations and secondary indexes. So we can support the full breadth of use cases you're used to building on our platform. Now another challenge we saw out in the industry was that most serverless databases force a hard trade-off. Either you want the benefit of being able to bring your cost down to 0, but then you pay the tax of a cold start delay should your application spin back up or you want the benefits of capacity can scale up very quickly, but then you're paying for pre-provision capacity and wasted costs. With Atlas, there is no trade-off. You can scale down to minimal usage and instantly scale up as your application needs without any pre-committed capacity. We also looked at the typical pricing models. And one of the biggest complaints we heard across our customer base was that with serverless models, the cost can run away from you. If your application spikes, the cost scales linearly alongside your usage. So we worked very hard in our architecture to enable a model where we -- our costs scale and give you discounts and benefits automatically and don't scale linearly with the workload that's being served. In fact, we discount up to 90% of all read usage on a daily basis without any preprovision capacity, no upfront commitment, a first in our industry. Serverless databases are also quite limited in terms of locking you in. And that shows up in a couple of forms. It could be that your business or your application requires cost predictability or the isolation of a dedicated cluster. Well, you can't often move from a serverless database to a dedicated database very quickly or easily. It requires a lot of rewriting or a complex migration. And of course, the prevalent popular serverless databases lock you into a single cloud provider. So your organization or your business demand the flexibility or a regulator demand the flexibility to move to another infrastructure. Again, you're rewriting your application. With Atlas, we give you complete freedom. We give you the ability to choose a serverless or dedicated model to move from one to another without any code changes. We give you the ability, of course, to deploy on AWS, GCP or Azure, any of our major cloud partners. So I'm extremely excited to announce that Atlas serverless is now generally available today for all of your production needs. Atlas serverless is a database without compromises. It gives you the full power of MongoDB, the deployment model and cloud flexibility or to use desire, usage-based discounting baked into the core pricing model and no scaling trade-offs. Now the core database is only 1 piece of our overall serverless strategy. In fact, many of the capabilities in the Atlas platform are in a consumption pay-as-you-go serverless model. Things like charts for visualization, our app services such as triggers, functions and our synchronization technology, our data lake engine as well as our data APIs. And you'll see more and more of this from us as we continue to invest in our platform going forward over the coming years. But serverless is happening up and down the entire application stack. We're seeing monitoring and observability CI/CD services, Functions as a Service, front-end development platforms and API services all being delivered in a model that's easy to use and get the complexity out of the way. And to this end, we are very excited to partner with Vercel. Vercel a leading front-end development platform that allows you to develop, preview and ship front-end applications end-to-end with ease. And I'd like to introduce the CEO and founder of Vercel, Guillermo Rauch, to tell you how we're working together.

Good morning, everyone. It's great to be here with you today in New York City. MongoDB's bet on the serverless in its movement is a great bet on where software development is headed. Vercel provides a serverless front-end infrastructure for building on the web. We enable developers to create at the moment of inspiration. Together, Vercel and MongoDB deliver on the promise of full stack serverless development. So why does Vercel pairs so well with MongoDB. Vercel is easy, scalable and fast. These are the core values of our front end platform. Let's start with easy, my personal favorite. Making things easy and creating a great developer experience has been the calling of my career. I first did this when I created an open sourced Mongo's, the leading client library for MongoDB in the JavaScript ecosystem. I then took that experience of simplifying the back end. I took it to the front end with a new challenge. And I created Next.js and Vercel. Next.js is the most popular framework in the react ecosystem. It turbocharges your application with server rendering automatic production optimizations. And the best part, it brings the full serverless development stack to your local machine. With over 2.5 million downloads every week, it has become the front end framework choice for initiatives of all sizes. Some of the most recognizable brands in the world like Dior and GitHub are reinventing themselves with Next.js. And this is happening all across the board. Companies like TikTok and Twitch are shipping some of the most incredible mobile web apps entirely built on Next.js. But Next.js, it's just 1 example of an exciting open source front-end ecosystem that Vercel contributes to and invests in whether it's a [indiscernible] or Next, there is no shortage of modern options to build with today. And fortunately, even after a developer chooses the right framework for their project, they might still struggle to ship high-quality software to production. Doing this today involves juggling all these different vendors, CDNs, clusters, functions, cashing infrastructure, Kubernetes. Vercel is the solution to this problem, allowing developers to go from get push to global application in 1 unified process. In practice, this means Vercel builds and runs our code, your front-end code base, natively in the cloud. So let's now talk about scalable. The promise of serverless that we always talk about is that you can scale up and then down to 0 as your traffic dictates with 0 effort on your side. So how do we actually deliver on this promise. We start with our stack and the choice of data platform. Obviously, it's MongoDB. When you deploy a modern framework like Next.js or Vercel. We take our code and we deployed a serverless functions that fetch from that data API. These functions are later used to serve render your pages. But we all know that scaling is not just throwing compute and money at the problem blindly, whether that compute is serverless or not. As visitors come in, Vercel intelligently renders and later reuses your pages, keeping the content in sync with your database. And what's best, this final layer is globally distributed at the edge. So, so far, we've made it easy to deploy but also automatically scalable and maximally efficient saving your resources. The result, a global application that's resilient and fast everywhere. With the built-in observability in real time to keep it so, keep it fast over time. So this all sounds great. It looks great. Look at that dashboard, has dark mode. It's beautiful. So how do I get all of this as a developer today? Today, I'm really proud to announce the official MongoDB integration for Vercel. Now MongoDB, Next.js and Vercel gives any developer the power of full stack serverless technology. So let's now watch a demo of how we build the starter app [ mongodb.vercel ] app, which you can visit now, and we will be deploying it in just a minute. Let's watch. I'm on Vercel dashboard and a click install the MongoDB integration. It takes me to Mongo. I connect my accounts, I provision my cluster. And then I don't spend any more time in the MongoDB console. This brings them back to Vercel where we build and deploy our application instantly. When the built completes, we're going to push your application to the global edge. Now [ mongodb.vercel ] app is live, Check it out and register your profile. Get started today with MongoDB and deployed to the global edge with Vercel. Visit vercel.link/mongodb to install our integration. Thank you.

Thank you, Guillermo. We're very excited to work with Vercel on pushing serverless in the industry further. Now as we take an idea from inception to a global platform, modern applications increase the demands we have in terms of features and capabilities as we all iterate, day in and day out. And so as we focus on adding broad workload support in our developer data platform, we want to make sure we pay attention to all the most critical needs of modern applications. Now the best way to walk you through this is to use an example. So if we're going to walk you through a hypothetical application we're working with today. We love the green theme at MongoDB, as you can tell by the colors all over this hall. And so our application is called Leafy. It's an innovative dual-sided marketplace that allows sellers to grow all the interesting plants that they have in their greenhouses and find customers worldwide and allows all of us to find the perfect plant for our homer offices. And of course, there's Leafy Corporation, the amazing marketplace business that enables this worldwide botanical commerce. So let's get started. First of all, in Leafy, I want my end users to easily find what they're looking for. Now the most ubiquitous way for us to interact with modern applications is through search. Now the challenge with search is that it adds a lot of complexity to your infrastructure. Not only do you have your core operational database perhaps driving your transactions, you also have to stand up a separate search cluster. The separate search cluster, of course, has to be kept in sync with your core operational data. So you end up installing an ETL pipeline and processing pipeline or a stream to reconcile and move data back and forth. And when we talk to customers, this extra integration, this multiple components in the architecture can drag down development by over 20% to 30%, which is ugly. And so a few years ago, we launched Atlas Search. It's the simplest way to enable rich and fast relevant space search in your modern applications and we've seen this product take off. Now as part of my Leafy experience, I don't want a search bar alone, I want to enable the classic filtered experience that you might see on the left now of an e-commerce side or a catalog system. I want to be able to have plants organized by size, by atmospheric conditions or the lighting conditions by which I'm able to display them. And so I'm excited in MongoDB 6.0 to release search faster, which makes it super easy to create these compelling searchable experience out of the box. We're also adding cross collection search, making it easier to search across multiple sets of data across a complex application. And support for embedded documents and arrays, bringing the full richness and power of the document model for the first time, bringing it forward into a search experience. To learn more about the flexibility and power that Atlas Search brings, take a look at the talk later from Keller Williams. Keller Williams is the world's largest real estate franchise. They've consolidated their architecture and simplified their development by using MongoDB Atlas with Atlas Search. I absolutely love the title of their talk, The Keller Williams Atlas story. So check it out. Now as I built my Leafy application, I started to realize the majority of my customers are actually on a mobile phone. My sellers, they've got all their plants in a greenhouse. They want to be able to easily choose inventory and update and list new products. My buyers, they want to buy plants on the go, maybe they're out and they see something in the wild and they say, "I need one of those." They want to be able to buy it on the spot. Now most modern applications are expected to have a mobile-first or mobile-only experience going forward in many cases. But of course, enabling this is quite complex. You need dedicated teams who manage the interaction between the persistence on the mobile device and the API and integration back to the cloud where all your core data sets. It doesn't sound that hard, right? In fact, it actually is. Dealing with the complexity of personalized data down to every device, dealing with conflict resolution the many edge cases. What happens if a data -- your user goes under a tunnel for a few minutes and losses connectivity in the API request is in the format that the client expected. How do you handle all of these different errors and edge cases? Well, you handle it by building a lot of extra code. And so we saw an opportunity to simplify this. A few years ago, we were excited to partner and acquire the team from Realm. Realm had built a novel mobile database that for many reasons, mirror why MongoDB is popular. It provides a much more natural way to work with data on mobile and edge devices than dealing with the native solutions embedded in the operating systems. And we've worked tirelessly over the last few years with our joint teams enabling Atlas devising, which get rid of all the custom code and the cobbling together of APIs to synchronize data across devices or to the cloud. So this year, we're excited to announce the GA of Flexible Sync, a new model for synchronization that allows you to define a query that can provide granular access control to which data gets to every single device. So you can provide a perfectly personalized experience with the security you need. We're also adding the preview of asymmetric sync. This is for right heavy workloads. Think about the connected car example Dev had brought up earlier. You want to collect a bunch of information across a bunch of cars, aggregated it back in the cloud for real-time analytics of what's happening in your business. We make that simple. And of course, we want to stay in tune with the latest languages that mobile developers prefer. And so this year, we released the Kotlin & Flutter SDKs, making it easier for developers to work in their language of choice. A great example of a customer that's leaned all in with Atlas Realm Sync, is Flowbase. Many of you in the room may not have heard of Flowbase, but they're an actually amazing business. They empower millions of SMBs in India to provide digital invoicing and accounting for their business. So you think about whether you're in a large city or you're in a village, everyone is now digital and accurate with all their information. The most exciting part about their story is they were able to replatform their whole customer experience in less than 6 months with MongoDB Atlas and Realm. So now my Leafy users have an amazing mobile experience. They can search the products they want. What's next? Well, it turns out in the dual side of marketplace, the buyers want to understand real-time information about the products they're searching for. They want to understand what the latest price was. Are they getting a good deal. They want to look at how many other people are looking at this particular plant. So there's a lot of competition. Now oftentimes, this type of experience has a very different pattern, a different query shape, a different set of data to power that real-time experience. And oftentimes, that adds complexity. Now as a seller in my marketplace, I want to understand the state of my business, how much revenue am I going to collect? What plants are most popular? How is my price history changing over time? What's my real-time traffic. Enabling this, again, drives more complexity into the infrastructure. Oftentimes, we see bolt-on solutions. We see an operational database, of course, bolted together with the streaming technology. Maybe a separate analytical database to handle the aggregations. You want to combine those results sets and then you've got to add a cash, maintain that consistency of that cash with multiple different systems. All of that is more cost, more trouble and more time spent away from innovating on my Leafy application. And so for years, we've been working on making sure that MongoDB has great capabilities for enabling these rich real-time experiences. But in the last year, we've got even further, we've added window functions that make it easier to track a running total or a moving average as part of an application. We've added new accumulators and operators like top, bottom, min, max, they push processing down to the database so you don't have to build extra code in your application to handle that logic. And we added Atlas analytic nodes, giving you the workload isolation to allow these complex analytical queries to be separated in your database cluster from our core transactions, so you don't interrupt the flow of revenue through your business. So from the thousands of customers that have deployed Atlas analytics nodes, we have 1 clear peer who feedback. They wanted the ability to tune the different hardware configurations independently, all within a unified cluster and database. And that's because oftentimes, your analytical query patterns. You have a different price performance profile than your transactional queries. Now you can scale them independently. You'll be able to spin up your analytics nodes, give them more hardware and memory to power those queries versus perhaps scaling out your operational queries to span a global user base. Appalling this type of rich experience requires a lot more than just tuning the hardware. And so we've been investing heavily in our core query engine, and we're excited to announce faster lookups. Oftentimes, analytics queries are joining multiple collections of data together, and you want the performance to allow those collections to return results in real time. To give you a sense of how much faster this is, I want to break it down a bit. Let's say you have a lookup that's matching a large volume of data, perhaps for Leafy Corporation, I want to look at a historical report of a year's worth of sales data. In this case, I could see a 2x improvement in query results for reporting. Let's say I want to shorten that window and I want to get a real-time view of last week's transactions, a much smaller number of matches. You can see 5 to 10x improvement in performance. But oftentimes, I may not even have an index supporting my particular query, perhaps it's an ad hoc query or exploratory resources are clearly generated by a particular user interaction. Those can be up to 100x or more faster in our new platform, fundamental investments that make it easier to build more performant, more rich applications. But that's not all. Oftentimes, an operational store is organized in roads. You're pulling back an entire document, updating a record based on a change to a customer's information or you're processing a transaction. And you're doing that with low latency and at mass concurrency. So we can serve millions of users worldwide. But oftentimes, that type of indexing is very different than what's required for an analytical query where you may be aggregating information across multiple records. In that case, you don't want to pull back every single record every time. So we're excited to announce column store indexes coming in MongoDB. This allows you to consolidate analytics queries in real time in your core operational database. And by scanning less data by not retrieving all the core documents and enabling faster aggregation and you can see a 15x improvement in analytical queries powering your real-time applications. Many of these innovations are behind the scenes. They don't show up on a feature list, but they're fundamental to the experiences you build day in and day out, and we're constantly pushing the ball forward. Now so to get back to my example, oftentimes, analytical information, powering an application is often time series data. And my example here, I'm looking at price history over time, and I want to be able to understand how those roll up over time and run interesting analysis around that. But that oftentimes forces customers to have to create and manage a separate time series database that's optimized for the cardinality and performance requirements of this unique type of data. So last year, we innovated and announced time series collections native in MongoDB. So you didn't need a separate time series database. You could power the rich capabilities of a document-oriented database but get the performance benefits and efficiency of a time series system all in one. And so for the past year, we've been investing heavily to keep up with the customer demand of time series collections. We've added charting support, densification and column our compression to make it more efficient. Gap filling. So if a sensor falls offline, we can automatically fill in the data and still serve the results set. And now in 6.0, we're excited to announce geo-indexing, full secondary indexing support and reperformance improvements for queries such as sort [ end point queries ]. One of my favorite examples of a great time series use case on our platform is J.B. Hunt. J.B. Hunt is one of the largest logistics companies in all the United States. They've leveraged our time series capabilities to manage sensor data from carriers all across the country. This allows them to make better short-term and long-term decisions and remove friction in the supply chain as much as possible. It's a great use case in logistics in time management that we've enabled in our platform. Now Atlas has become a platform that's much more than just an operational database. It's a unified platform for real-time in-app analytics. We enabled Atlas analytic node tiers to improve performance and price performance tuning of your cluster. We've added faster lookup so you can get better join performance in our database. We've added column store indexes coming soon to give you analytical performance and an operational database and time series collections, removing the complexity and need for more niche databases in your architecture. This Leafy application is getting quite broad. Now I've talked about the sellers. I've talked about the buyers and their experience in this new application. But what about Leafy Corporation itself. And we all know it takes a lot more than just the development teams building the application to take an idea and make it a global reality. In fact, there are many other stakeholders in an organization. There may be analysts that create reports to help product managers or executives make decisions about what direction to take the business in. You may have data scientists who can't even do any work without massive amounts of data and to be able to understand it, run models to help better decision-making across the organization in automation across the organization. And of course, you have data engineers who have to massage, integrate and manage the data across the entire organization. Now many of these folks are looking at data in different patterns. Most commonly, you want to look at historical data. In my case, I want to look at yearly sales. I want to look at last year sales versus the year before or this year versus last year to get a better sense of the trajectory of my business, so I can plan better. I want to then take that data and start thinking about the future, what geography should I enter, what plants are most popular that I want to start growing now and advise my sellers to start creating. But as we look at these common use cases, we found that something is missing. We found that it's really hard to get a real-time view of what's happening in the business for these other stakeholders. And as Dev mentioned, almost every business is going digital. So the ability to react in real time to what's happening to the business can mean the difference between thriving and failing for many organizations. And that mission-critical state of the business, the real-time information of what's happening lives in your applications. So we think there's an opportunity to make this much simpler. Imagine if the business could have a real-time view of transactions, of users, of traffic, of sales, to get a glimpse of my business and react faster to changes in market dynamics. This is a great thought. In many organizations, this is impossible. And for those who can, it's extremely complicated. You need multiple data sources, streaming them together, multiple ETL processes built by data engineers, integrating all these different moving parts, and it's very hard to get a consistent view that's accurate. And then people making those decisions, the analysts, the data scientists rolling up that information so that the business can pivot and move in the right direction, oftentimes are working with stale data because all of that flows downstream and a batch process to a data warehouse that's best to day old, if not many times, weeks or months. We think there's a fundamentally better way to do this. And so to walk you through the new capabilities we're offering to serve the entire data life cycle, I'm really excited to welcome Wisdom Omuya, Director of Engineering at MongoDB to walk us through the new announcements. Wisdom, all yours.

Thanks, Sahir. Let's now walk through a few new capabilities of Atlas to help Leafy Corp meet their needs. First, let's consider the requirement of maintaining a historical view of the business so that over time, we can understand what's happened and identify trends for the long term. To meet this requirement, I'm pleased to announce Atlas Data Lake storage. Atlas Data Lake allows Leafy Corp to store large amounts of historical Atlas cluster data without having to set up complex ETL workloads to see how that works, let's head over to Atlas. I'm going to be creating a data lake pipeline for the plant collection shown here. Our Data Lake will allow us or historical data sets from this collection. Over here in Data Lake, I'm going to create a Data Lake pipeline. We'll use the Leafy Corp cluster and the Leafy Corp database within that and finally, our plant collection. Now the business analysts at Leafy Corp want to review this data on a daily basis. So I'll duly extract my data sets on that cadence. And then to optimize performance, we can specify regularly queried fields. I'm going to use the plant family here. And finally, since this is plants data, nothing particularly sensitive. I don't need to filter out any of the fields in my collection, so I can just hit finish, and there you have it. The Atlas Data Lake pipeline I just created is what you see here. Now what we've just done is set up an automated pipeline to extract consistent snapshots from our Atlas cluster and store those in a fashion optimized for performing analytics. There's no ETL tooling required here, no plumbing to manage. Okay. So I now have my data set. How do I query it. To query it, I need to use Atlas Data Federation. For the demo today, I've created this plants pipeline collection here that has a number of data sets as you can see, that have already been ingested. Now it's important to note that each data set here is consistent as of the point in time shown in the data set name. So to carry my data sets, I'm now also pleased to announce Atlas Data Federation. Our federated query engine that allows you to query across a variety of data sources, including Atlas Data Lake. So let's go ahead and do that. I'm going to create a federated database. The first one I'll do here is rename my instance to Leafy Lake and I'll rename my database, my logical database here, Leafy Plants. Now as you might imagine, we're going to create more and more snapshots of our data over time, and Leafy Corp needs a way to access all of this data, the instant they're available. Atlas Data Federation supports a wildcard collection functionality that allows us to do just that. So I'm going to select that functionality here and then save my Federation instance. Awesome. So let's now do some querying of this data I'm going to connect this time with the MongoDB Shell. So let's copy the URI there and then hand over to my terminal. Awesome. When I show my databases, you can see I've got the Leafy Plants logical database that I just created. And when I show collections, you can see I now have all of the ingested data sets from my Atlas Data Lake. So now we can do some analysis of our data. Let's run in aggregation to figure out, for instance, how many plants are in each plant family in my 4th of June inventory. Excellent. So what we've just seen is the ability of Data Lake to store and analyze a lot of historical data for the Leafy Corp organization with just a few clicks. If you take a step back, what we've done is a couple of really impressive things. First, we've seen the ability to set up a pipeline to capture consistent snapshots of our Atlas cluster data; second, the ability to store that data in a storage service that's optimized for performing analytics, both of these using Atlas Data Lake; and finally, with data federation, the ability to persist that data and query it in a way that's optimized for our use case. Now data federation not only allows you to query historical data. It also gives you a great real-time view of your data. So let's see how that works back here in Atlas. Now as you can see, the Leafy Corp application is actually powered by multiple Atlas clusters here. And Leafy Corp needs a full picture of their business. In order to do this, they need to be able to query across each of these clusters that you see here. Furthermore, Leafy Corp also has S3 data that's enriched with plant characteristics, data that served externally but also used to enrich the plants collection that you saw earlier. This information helps with your buyers and your sellers to make good decisions in your marketplace. So the question now is how do we support Leafy Corp to query across Atlas clusters and S3? Well, we previously have had to carry multiple databases and then combined the results ourselves or we'd have had to perform some sort of ETL to move the data around. This is no longer necessary. Atlas Data Federation provides a single unified quarry endpoint that allows you to access all of your data across a variety of data sources. So let's see how that works over here in Data Federation. So I have this Leafy federation instance. And if I look through the configuration, you can see that data federation allows you to source data from Atlas clusters, S3 data lake as we saw earlier, and even HDP data sources, if you want. Now on the right here, I have a federated plants logical collection. This collection comprises data from my plants collection, as you saw earlier, as well as S3 data. This S3 data contains things like when the plant blooms and its sunlight requirements. And reaching our planned collection with this S3 data helps our buyers in the marketplace to make informed decisions for their purchases. So let's now connect to this Data Lake. This time, I'll use MongoDB Compass, which is MongoDB's Gui Tool. Over here in Compass, you can see I have all of my collection data from the Federated plants collection here. This is the same way I would browse through a regular MongoDB collection. Now it's important to appreciate that data Federation allows you to query transform and persist your data across a variety of data sources. So let's look at aggregations here to see how that works. Now at the top here, you can see I've got data from my S3 bucket as well as from my Atlas cluster. Now the first thing I'm doing here is using a dollar group to consolidate the information from both of these sources under a single field here that's called plant details. And if I expand that, you can see I now have information for each plant from across S3 as well as my Atlas cluster. Next, I'm using dollar replace route to unnest that plant details field into the top level of my document. So we've now seen how to query, we've seen how to transform, but how do we persist. Well, at Leafy Corp, the business analyst team requires that all analysis information be persisted as BRK files on S3. Data Federation meets this requirement by adding support for doing dollar out to S3 operations, and that's what you see down here. This allows us to share the enriched data with a wide audience both within [indiscernible] and externally. So let's do just that. Excellent. So it's really important to appreciate that this is something that previously would have required a lot of time and hard work from a data engineer, but now with Atlas Date Federation I can accomplish it with just a few simple clicks. In fact, one of our customers, [ Klim Base ], is giving a talk later on today on how they use dollar out and data federation to support some of their reporting needs. Or those of you who are watching online, you can watch that section on demand. So we've shown you how a single data endpoint is accessible from familiar MongoDB tools like MongoDB Shell and MongoDB Compass. But how about other folks who use different tools for access in their data. Well, at MongoDB, we want to meet our customers where they're at. And to walk us through how we do that, I'm pleased to introduce Alexi Antonino, Product Manager for Atlas SQL to walk you through additional capabilities of our developer data platform. Alexi?

Thank you, Wisdom. Leafy Corp has multiple teams with varied responsibilities and they all need quick access to real-time and historical data to make business decisions. One way to achieve this is with Atlas charts because Atlas charts allows you to natively visualizes MongoDB data in a way that takes full advantage of the richness of the data document model. But -- and Atlas charts now allows you to easily share and embed full dashboards within your applications. But at Leafy Corp, we have multiple teams already using the SQL-based tool, Tableau for their reporting needs. To help these folks, I'm very pleased to announce the preview of the new Atlas SQL interface. This new SQL interface allows users to gain access to live and historical data even when using their SQL tools. Now some of you may be familiar with the BI Connector is a way to gain SQL access to MongoDB, but this SQL interface provides an enhanced experience, 1 with a richer dialect and tighter integrations with SQL tools. Let me show you how easy it is to connect. Here I am in Atlas within data federation, and I can use that same federated database that Wisdom just created and configured. Since it's bringing together and combining all of that great Leafy plant marketplace data I would need if I was a Leafy CorP analyst, tasked with building dashboard views of the business. So as an analyst, all I need to do is grab some connection information whether I connect using the JDBC driver or the new Tableau named Connector, which I plan to use, I just copy this connection information and power up Tableau. Once since Tableau, I now see my new Atlas MongoDB Connector in the Tableau Connect menu. So I just simply entering my connection criteria starting with that copied in URI from Atlas, and I'm ready to begin querying. Here's my Leafy plant data, and it's now coming in as tables with columns and rose. But what's really interesting is how the nested document data and the listing array data is displayed. See this stringified format. So here, we have the buyer nested document and this listing detail array. And that stringified format that you see there is by design because it allows me as the analysts to profile that data, see its content and then decide how it should be modeled or transformed to build visualizations. And I can do that transformation using SQL because Atlas SQL exposes Tableau's native features like this custom SQL Editor. So I can use SQL along with dot notation to flatten that buyer nested object. So here I am, and I flattened age, gender and bio satisfaction score to create distinct columns for those fields. Then I'm using a SQL case expression to put those buyer ages into buckets or age range groupings for a better, more impactful visualization. And once I'm done, transforming this data and readying this data, I'm ready to start building and enjoying and exploring these visuals.

Alexi, that's awesome. You mentioned that Atlas SQL is available today. Can you give us a sense of what's coming next?

Yes. Actually, we have plans to support an ODBC driver and connectors for Microsoft Power BI as well as Google Looker. And those are just the first of many SQL interface integrations to come. So stay tuned.

Thank you, Alexi. Amazing demo. So as you heard from Wisdom and Alexi, Atlas is now a platform that can serve the entire data life cycle. Atlas Data Lake, which allows you to tier your data or store large volumes of historical data for analytics cost effectively. Atlas Data Federation, which allows you to combine multiple sources of data together easily, perhaps merging historical and real-time data to get a full picture of your business. Atlas charts document native visualizations, not only in dashboards, but also something you can embed live in your application experience. And of course, the Atlas SQL interface, which gives you the ability to integrate and empower all of those in your organization with the tools they use every day. So as I step back, my Leafy platform has evolved into a really powerful application. It's serving a worldwide community of plant enthusiasts, creating amazing commerce across the world. But to take this app to production and scale it over time, it's not -- it's much more than features that I need. I need scalability, resilience, performance and global reach. So I can make sure my end users are always finding what they need easily and in a performant way. Over the past 6 years, we've brought on 35,000 organizations onto the Atlas platform, representing 100 different countries. And they're deploying in 95-plus global regions across our cloud providers and partners, AWS, GCP and Azure. There's no better way to express the power and scalability of the Atlas platform globally than to have a customer explain it themselves. And one of the most interesting businesses on the platform today is Avalara. They help companies of all sizes manage the massive complexity associated with the global tax regime. So I'm excited to welcome John Jemsek, VP of Engineering from Avalara to tell their story.

All right. Thank you, Sahir. Good morning. Who here likes taxes? They are my people right there. All right. So that's probably not the question you were expecting me to ask at a technology conference. But take a moment and imagine all of the financial transactions happening right now all across the world. That is a complex tax problem to solve.