Datadog, Inc. (DDOG) Earnings Call Transcript & Summary

Hi, everyone. I'm Olivier Pomel, CEO and Co-Founder at Datadog. I want to thank you all for joining us today. This is our second year running Dash virtually. And while I do miss meeting you all in person, I'm excited to see that the online format makes it possible for many more of you to join from anywhere in the world. Now the last 18 months have brought a lot of change to the world with many of our businesses moving online and transitioning to working from home. But one thing that hasn't changed is the pace of new launches, features and improvements from our teams at Datadog. We have launched over 300 features of integrations since we saw you at Dash last year, which is almost one new feature every single day. And we'll have more to come this morning, new silos, new use cases, new data sets and new ways to use that data. We have a very full agenda for you. But Dash isn't only about new announcements from us. It's an opportunity for all of you to learn from each other. In addition to the Dash Keynote, we have 2 full days of panels and sessions where members of our community will share how they're addressing scale, speed and security in the organizations. After the sessions, you will be able to stop by our virtual export floor to see demos of our latest announcements as well as to meet with our partners and sponsors. And you can also get hands-on and learn about all these new additions in the workshops we are running throughout the week. If you haven't had a chance to sign up, you can still join. And all of the process from workshops will be donated to PowerToFly, BreakLine and FlockJay, all organizations that invest in bringing more diversity to the tech industry. Thank you again for joining us today. We're all very excited to see what's coming next. And to kick us off, I'd like to invite my Co-Founder and CTO, Alexis.

Thanks, Olivier. Datadog's mission is to provide the visibility and data you need about the services and the infrastructure that power your business for many of you that has meant helping you understand your infrastructure as you have migrated to the cloud or containerize your workloads. In this dynamic world, our ability to focus on services instead of individual hosts or resources has led to auto-scale workloads and shift to the cloud to take advantage of it. But many of you also have workloads that live at the edge of the cloud, everything from IoT devices like digital billboards, medical equipment, dairy farms to on-premises data centers and branch locations. Whether powered entirely within your data center or in a hybrid environment, these workloads are key and require the same level of visibility as your cloud workloads. But these key systems take more than just an API call to manage, and they challenge us to think about visibility in a different way. To share more about how we're thinking about these workloads, I'd like to hand it over to Natalie Altman.

We built Datadog first as a product to help our customers gain visibility into their infrastructure. Almost all of our original customers were using the product to monitor their cloud environments. Whether they were cloud-native, early adopters or had recently made cloud migrations, existing tools didn't effectively support monitoring these dynamic environments. So they turn to us for help. But the more customers we spoke to, the more we learned how many of them were still maintaining physical environments and how many of them always will. Whether it's because they need to cut as much critical compute time down at the edge as possible or because in-house operations at such large scale are better optimized with network experts with institutional knowledge, or maybe it's because they need to be physically close to their customers in retail stores and supermarkets or corporate offices where IT ops need visibility into access points and the WiFi. Ultimately, your applications communicate across this equipment through gateways across the LAN over the WAN between switches and routers leaf to spine, distribution to core hub bespoke with BGP or OSPF, pass an app through direct connect to your cloud infrastructure and beyond to your customers. Slow network communication can occur at any point along the way, which means troubleshooting an issue can span across the NOC and DC ops and app and DevOps teams who often don't rely on the same troubleshooting tools, making isolating the root cause all the more difficult and time-consuming. This is why today, we're thrilled to announce Network Device Monitoring. With Network Device Monitoring, we're breaking down yet one more silo: for the network engineer who needs visibility into their rack and stack network infrastructure, like their servers, routers, switches and firewalls in the data center over which application traffic flows, and the access points in campus and branch office networks to join the fold with other app, security and DevOps engineers already in Datadog. Instead of stepping through runbooks after an incident, trying to isolate where of a series of issues may have occurred, network engineers can collect to report and alert on important telemetry data, like saturated switch ports or drops in connections to a firewall or circuit failures. And with integrations with ticketing and alerting tools like PagerDuty and ServiceNow, you can make sure all the right stakeholders are alerted when customer-impacting issues arise. Or with machine learning algorithms, like anomaly an outlier detection and forecasting, detects and fix an issue before customers ever experience any change at all. With visibility into your network infrastructure alongside the applications that communicate across it, teams can now move even more quickly to isolate where an issue may be occurring no matter how large their environment. Wayfair, a leader in the e-commerce industry for furniture and home goods, has a large complex network infrastructure composed of thousands of devices. With hubs at their data center core and many remote in campus sites, ensuring that everything from core routing and switching gear, devices at the edge and remote gear are healthy, it's critical for their customers to have a good experience. But when there's a report of latency on the network, their NOC team is on speed dial. As an existing APM customer, they're already a step ahead at identifying where the source of the latency may be coming from. Their app teams have the visibility to identify that one specific application that they know makes its way across the link to their data center is failing. Let's take a look at an example of how they might isolate where along the network is causing this issue with Network Device Monitoring. Say we have 2 data centers, 1 in New York at Datadog headquarters and 1 in Paris for our EU engineering teams. We want to start just by looking for the links where the application is making requests. We know there are connections that go through our data centers, so we can start at the edge and check for whether we're seeing extended traffic or whether a circuit may be overloaded. We can filter to see only the devices from our data center in Paris and then group by device type, so we can isolate the performance from the ones at the edge versus those at the core. Let's check in on the edge first. If there are connectivity issues here or changes in traffic patterns, it may be the source of why we're not getting any data. Immediately, we see that links from the edge are overly saturated, which must be where the latency is coming from. In just a few clicks, we've isolated a network issue across our app teams through the data center and can create a ticket for our ops team to go in and solve the problem. What I've shown you today is how customers like Wayfair can gain deep visibility into their entire network infrastructure using Network Device Monitoring. They're able to isolate issues between infrastructure and application teams, allowing them to move faster and more effectively, spending less time trying to isolate where the issue may have occurred and more fixing the problems as they arise and finally, make sure they're alerted before customers report these problems in the future with machine learning-based monitors and integrations into all the leading ticketing tools. Network Device Monitoring is generally available today and can be enabled right from the Datadog agent. Please reach out if you have any questions.

Thanks, Natalie. Keep breaking down those silos. Network Device Monitoring focuses on the routers, the switches and other networking hardware that connect your applications and services, allowing you to understand every part of your infrastructure, bridging the monitoring divide between your on-premises environments and the cloud. Now that we've connected those silos, our next announcement moves up a layer to focus on what runs on top of your network, your applications. Because that's one thing every modern application has in common, whether it's a monolith or a microservice. It talks over the network to other applications and services. This network traffic offers us a great vantage point to understand the health of your services and observe how they interact. For that, I'd like to pass the mic to Stephen to share a bit more about our latest addition to infrastructure monitoring.

Your architectures continue to get more complex, spanning containers, serverless functions and even multiple clouds. Your teams choose to build on these new platforms, ship features faster than ever. You need to know if your services are reachable, how much traffic they have and if they're throwing errors. The Google SRE handbook defines these metrics as the golden signals that every customer-focused team needs. Both applications changing so quickly, legacy code running in the background and new services being added all the time, how do you ensure you don't have blind spots? How do you understand your service dependencies? And how do you know that all of your services are healthy? I'm excited to announce that you can have ubiquitous immediate visibility into the health of all of your back-end services without changing a single line of code or redeploying your applications. We call it Universal Service Monitoring, and it's based on the eBPF Linux technology that guarantee security and low overhead. By changing a single configuration in the Datadog agent, one SRE or infrastructure engineer can roll out Universal Service Monitoring for hundreds of applications at once and instantly build a complete inventory of their services, their interdependencies, health metrics and deployment details. And this integrates seamlessly with your infrastructure monitoring, custom dashboards and Datadog monitors. Let's take a look at Universal Service Monitoring in action. I'm going to show you how to have observability into your entire constellation of services without needing to modify or redeploy them. This is an application that one team owns. My team is responsible for ensuring the health of dozens of applications. After I set up Universal Service Monitoring, I can be alerted to issues with high latency or errors from any back-end servicer team and tie these back to my organization's SLOs to get an accurate picture of our health as a business. In one place, I have an inventory of all of my services, a visualization of their upstream and downstream dependencies as well as golden signal metrics for each one. With this information, I can decide whether to scale up, roll back or bring in a specific team to help. Universal Service Monitoring is the best way to give broad visibility across all of your applications and to detect any issue as it unfolds from day 1. Once you know there's a problem based on your golden signals, you can get a more granular view of each full end-to-end request by setting up distributed tracing to then pinpoint a line of code or method-level bottlenecks, enable the Datadog continuous profiler. What I've shown you today is that Universal Service Monitoring is enabled without touching a single line of code or redeploying any of your applications. It gives you golden signals from all of your services that you can use to power monitors, dashboards and SLOs. And it helps you localize issues, starting from your full inventory of services and interdependencies. We're excited to take another step towards breaking down silos between Dev and Ops teams with Universal Service Monitoring. Sign-up today for private beta access.

Thanks, Stephen. Observability for every service everywhere without needing to change a single line of code or redeploy anything, how much easier can it get? So far today, we focus mostly on infrastructure and the network your applications depend on. Now let's shift gears and talk about the data they generate. Datadog's logging without limits has changed the game for how you think about your logging data by enabling you to ship all your log data but only retain what you need immediately at your fingertips. However, we heard a few challenges from some of you. For instance, the log data you wish to retain can often be very voluminous and costly to keep long term. And you want to reuse that data in other systems outside of Datadog for observability and business needs. And finally, some data sets may be too sensitive to store in third-party systems. We have listened. Michael Whetten will join us and share some exciting new offerings that will allow you to bring these large, sensitive and multipurpose data sets into Datadog.

Last year, we had a customer come to us that needed to run an exhaustive investigation. We cover more than a year's worth of their logs, over 30 petabytes of data, spanning multiple applications, services and systems. A single pane of glass would be critical to keep tight coordination between their DevOps, security and legal teams, also to help them correlate system forensics from metrics with application and user behaviors from their logs. They weren't sure how long the investigation would take only that it needed to be comprehensive, so it could take months. This was a new type of investigation for us and had opened our eyes that there's more that we could be doing to help our customers. You see they used our hot storage indexes that provided sub-second queries over their critical operational logs, and they use these for monitoring and troubleshooting production environments. But it was too hot to affordably keep these logs for more than 30 days. They also sent all their logs to external archives, rehydrating logs back into indexes versus smaller historical investigations. But at this scale, rehydration would be cumbersome. What they really needed was an online warehousing solution for their observability data, a solution where they can afford to keep massive amounts of data for years at a time, where they can run complex investigations or look at extra-high-cardinality trends on historical data. Queries could be slower because completeness and flexibility was more important than speed, but it still needed to be interactive to not bottleneck the investigation. So we went to work. And today, it is my pleasure to announce a brand-new, always-on observability warehousing solution that we're calling Online Archives. Online Archives is a new storage and query tier for Datadog log management that fits between indexing and external archives. With Online Archives, you retain all of your logs fully searchable for 15 months at a fraction of the cost of hot indexes. Let's see how it works. An online archive can be added to any existing index route, which usually represents one of your engineering teams or departments. Simply open the index configuration and enable Online Archives. All logs are kept in Online Archives for a default of 15 months. Here, we opened the log explorer and initiated a search 3 months back. Immediately, we see there are no logs in our index because we only keep them hot for 15 days. Switching from an index view to a historical investigation is as easy as flipping a toggle and all of the navigation and analytics features by transactions, patterns and aggregations still work. Inquiries can be used for evidence building and correlation in notebooks and dashboards. We can access the historical log files with all their tag context and even correlate the raw log with historic metrics data, which is also retained for 15 months. Online Archives is the next evolution of logging without limits, enabling your teams to keep more data interactive for longer periods of time without sacrificing the budget. Online Archives are currently in limited availability. So reach out to your Datadog representative to learn how you can get started. As we just saw with Online Archives, Datadog helps you to establish a culture of observability by promoting collaboration across teams to build healthy, secure and performing applications. But observability in a large company can be difficult. This is because you have hundreds of teams with diverse needs, thousands of applications, different levels of maturity and an ecosystem of legacy tools and vendors that you need to coordinate. All this complexity leads to runaway costs as data volumes grow, difficult investigations because of missing or inconsistent data and compliance and security risks as PII leaks throughout your infrastructure. Data silos result that are painful to migrate or lock you into a single expensive vendor. So today, I'm thrilled to announce Datadog will introduce Observability Pipelines, an on-premise organization scale data management solution. It runs on your infrastructure, whether that's local hardware or in the cloud, which allows you to centralize the collection of all your metrics, traces and logs into a single place. And then you can enforce a unified strategy around processing, redaction and sampling on all of your company data before it ever leaves your systems. You can send the processed data to whichever destinations your team requires depending on the use case. Most importantly, though, is that even though Observability Pipelines run on your infrastructure, you can figure, monitor and manage your sources, pipelines and destinations to Datadog's web interface, greatly expanding access and control to all of your teams for all of your company data. Observability Pipelines are built on top of Datadog's popular open source project, Vector, a framework for building highly performing streaming pipelines. Vector already has millions of downloads and a community made up of some of the largest companies in the world. Datadog Observability Pipelines are a game-changer for large companies that need to take control of their data and require choice in the face of growing complexity. You can get started with Vector today and reach out to a Datadog representative if you need help with Vector or need to learn how you can get early access to Observability Pipelines. Thank you.

Thanks, Michael. Online Archives enable even larger data sets in Datadog log management. And with Observability Pipelines, Datadog becomes a hub that gives you control and freedom over your data as you transform it, enrich it and route it to any destination. I can't wait to see what you make out of these new features. So far, we've talked about infrastructure, the applications it powers in managing large and sensitive data sets. Now let's talk about your experience in building and developing those applications. At Datadog, developers are the core of what we do, from monitoring the performance of thousands of microservices using APM to identifying the most obscure crashes across web, mobile and back-end applications where they are tracking. We know a lot of developers rely on us to deliver better, faster and more robust software. But there's more to observability than just applications in production. Applications don't just appear in production. There's a bit of a journey to get there, and that's the focus of the next few announcements. Starting with our new integration with GitHub, allowing you to correlate observability data from your applications and services with Git metadata. It lets you see the source code related to an error in one of your apps. It lets you quickly know who pushed new code during the latest release to production. This new integration helps bridge the gap between development and operations. But that's not where our integration into source control and developer experience end. We have more to say. And for that, I'd like to invite Borja to share a bit more about how we're providing visibility into the journey your code takes from development to production.

Today, with Datadog, engineering organizations get a sophisticated and rich understanding of their applications and infrastructure in their live environments. However, those live environments only capture the final destination of the software development life cycle. What about everything before? Continuous integration, or CI, as it's commonly referred to, is the beating heart of every modern organization's development process. But up until now, teams have lacked visibility into these critical systems. Any efforts to get visibility into CI required complex and custom in-house solutions that are brittle, require constant development and maintenance and are usually hyper-specific to an individual team's requirements. So the same solution must be built multiple times for different parts of the organization. And that is why we are excited to announce today the general availability of CI Visibility. This product represents a major step towards shift left observability and the first of many milestones in making Datadog your single observability platform for the entire software development life cycle from commit, all the way to production. CI Visibility is not just about bringing in raw test and pipeline data. We analyze and surface valuable insights and can automatically notify developers when a regression takes place. We have interviewed dozens of the most advanced software engineering organizations to distill a set of best practices when it comes to CI. With CI visibility, we have productized these best practices into a new product, enabling access to the cutting-edge CI tooling seen in the highest-performing engineering organizations around the world. Our CI Visibility product is divided into 2 distinct categories of functionality: pipeline visibility and testing visibility. Each bring their own unique view and insight into the 2 major facets of CI. Pipeline visibility captures detailed information about your pipelines and stages and jobs as they are being executed. Here, we provide you with a single pane of glass for pipeline data across all of your CI providers. We identify areas within your workflow to improve performance and reliability. And all of this data is available for you to create tailored dashboards and monitors for each and every team. Taking a look at a pipeline execution in CI Visibility, we represent each execution as a trace. The root span is the pipeline, and you can see that various stages and their jobs represented a spans underneath. The flame graph makes it easy to identify exactly why the pipeline failed and which stages or jobs are contributing to performance bottlenecks. We also ship this preset dashboard for pipelines, allowing you to provide custom views for each of your product development teams. This ensures these teams have a tailored dashboard to get quick feedback into their builds and pipelines, exact CI resource consumption as well as help troubleshoot any regression or identify areas for optimization. We are happy to announce that we support Jenkins and have partnered with many of the most popular players in the CI/CD space to provide native integrations with GitLab, GitHub Actions, CircleCI and Buildkite, and support for more providers will be coming soon. When it comes to testing visibility, we enable you to manage your flaky test, provide an APM-like end-to-end visibility for every test execution and allow you to track the historical performance of your tests to identify potential regressions. The product has been carefully created with the developer in mind, focused around the constructs you work with daily such as repositories, branches and comments. When it comes to flaky tests, we know that they can be extremely frustrating for the developer experience and, if left unchecked, can be quite harmful to the engineering culture and cause developers to slowly lose confidence in their testing. Datadog will automatically identify flaky test in your code base, along with information to help you triage this test and the ability to manage them. You can see information like the time and commit in which the test first flaked and its flaky rate since then. And Datadog will show you which commit was the one that either introduced the flaky test or involved a co-change that made the test flaky. We provide instrumentation for Java.net, Node.js, JavaScript, Swift, Python and Ruby. Alternatively, development teams can also upload JUnit test reports to send their test data to Datadog. And support for even more languages and testing frameworks will be coming soon. We think CI Visibility will quickly become an indispensable tool for developers and engineering organizations to better understand, monitor and troubleshoot the full end-to-end software development life cycle. CI Visibility is available today. If you are interested in learning more, we have a CI booth where you can stop by to see an in-depth demo and a workshop later in the week where we'll show you how to get started and do a deep dive into the full product capabilities.

Thanks, Borja. Helping teams ship software faster and with confidence has been the end game since day 1 at Datadog. It's been awesome to see the impact of bringing observability early in the process has offered our own teams at Datadog. I really look forward to seeing the impact it has on yours. But there's more. Once you get your applications to production, a common challenge you run up against is a challenge of understanding the performance of your databases. I know I've spent countless hours troubleshooting a poorly performing query and its impact on applications. Is the data set too large? Or is an index missing? Databases have been opaque for far too long, and that's why we're so excited when we launched Database Monitoring this summer, enabling developers and DBAs to better understand the impact of the database and queries on their services. To share more about Database Monitoring, Jason joins us next.

Database-related incidents are among the most painful for developers, SREs and database administrators. With the advent of microservices, application stacks have evolved to maintain separation and allow each other to degrade gracefully with minimal impact on one another. However, databases are still typically a shared resource, meaning that a fall-off from a single database instance going down can be felt by every engineering team and ultimately, your end users and customers. The effects of query performance degradation can bubble up all the way across the stack, impacting every consuming app along the way and, worse yet, may even result in data loss. Despite the fact that databases are so perilous, historically, they have been a black box for developers as they require unique knowledge and tools to effectively troubleshoot and optimize. Every company organizes their database ops differently, but it's always a shared effort between developers, SREs and DBAs working together to answer the question, what is wrong with my database? Do I need to optimize a query in an application, update the database configuration, scale my infrastructure? It's a hard question to answer, but it doesn't have to be. And this is exactly why we built Database Monitoring. Datadog's Database Monitoring product, DBM, streamlines collaboration between developers, SREs and database administrators, and everyone else relies on the database by consolidating all of your key database telemetry into a single view. Your teams can now spend less time finger-pointing and more time delivering high-performance experiences by seamlessly correlating the database performance metrics that we've always collected with new historical query performance metrics, explain plans and more using the tools that they already know. DBM lets developers easily identify which queries are their worst offenders using performance metrics that are familiar from the application space like latency and query volume with the added context from the database's perspective, such as the number of rows updated or returned. DBM also leverages your existing database tagging to make it easy to quickly scope down to address the queries you care about and answer questions like, what's used around this query, and in what data center? It also brings in new database telemetry, including explain plans, that provide deep visibility into the inner workings of the database, allowing you to discover the root cause of performance issues and determine what and where to optimize first. Combined, these allow anyone with any level of database expertise to understand where operational bottlenecks happen on a database and how to remediate them. Whether that means optimizing the query structure in a calling application to reduce the number of rows scanned and returned, updating the database tables to add indexes and avoid full-table scans or scaling the underlying infrastructure to better accommodate the CPU or disk loads, Database Monitoring can help you answer these questions. DBM currently supports Postgres and MySQL in any environment, regardless of whether the database is self-hosted or managed by a cloud service like RDS and Aurora, Cloud SQL or Azure. And today, we're happy to announce a beta program for Microsoft SQL Server as well. Alert on trends using query performance metrics, optimize latency using explain plans and query samples and slice and dice using tags to find crucial queries with Datadog's Database Monitoring. Get started today with the same Datadog agent that you already know, starting at $70 per host.

Thanks, Jason. Database Monitoring, troubleshooting and understanding your data tier has never been this easy. I wish I had this 10 years ago, but databases are just one of the many integrations Datadog offers with over 500 integrations. We've made it easy to collect data about every application in your stack from containers, cloud, CDNs, collaboration or ticketing systems. We connect to everything up and down your stack, allowing you to monitor, alert on and correlate data across the disparate systems that power your business. Until now, those integrations are mostly focused on collecting or exporting data. But what's most important about data is how you use it, how it becomes part of your day-to-day workflows. Up next, Ana Wishnoff will join us to share what the next generation of integrations with Datadog looks like and how you can build your own today.

Thanks, Alexis, and hi, everyone. Today, I'm delighted to announce our initial release of Datadog Apps. Have you ever needed to jump between your Datadog dashboards and other tools? Maybe you switched tabs to trigger an action in another app based on your Datadog monitoring insights. For example, maybe you needed to scale up a Kubernetes deployment. You may have jumped from Datadog to an external data set for comparisons such as checking your cloud efficiency score, or perhaps you have found yourself exporting data to build a custom visualization that wasn't supported by Datadog's out-of-the-box widgets, a box and whisker chart, for instance. Regardless of what you're trying to achieve, jumping back and forth between Datadog and other tools can be time-consuming and oftentimes disruptive to you and your team's workflow. This is why we're so excited to launch Datadog Apps. Apps will provide you with the external actions, visualizations and context that you need right inside of Datadog. You can now blend the Datadog platform with other applications to achieve simple and straightforward workflows. Simply put, Apps allow you to extend the functionality of the Datadog UI. In this first release, we've worked closely with a number of our technology partners to integrate external tools into Datadog dashboards. With new visualizations, interactive overlays and menu items, these custom dashboard widgets provide you with external functionality that feels native inside of Datadog. We'd like to thank our launch partners who have built apps that are available for you to install today. These apps are each specifically designed to make you more productive in Datadog. The Embrace app, for example, allows you to view insights into mobile app crashes that you can't find anywhere else right inside of Datadog. You can then use these insights to help resolve incidents, which you can do with the PagerDuty app that I'll be diving into next. So PagerDuty saw Datadog Apps as an opportunity to streamline escalation workflows and create a more seamless incident management experience by providing the context and actions that you need within your Datadog dashboards. Let's take a look. First, I'm going to drag this widget right on to our dashboard. Here, you can see a bird's eye view of all of your impacted business services in a list organized by priority with a summary of any related services that may also be affected. Over here is a detailed list of incidents that are currently active for your team. You can take action right in this app to either acknowledge or resolve any of these incidents without needing to go into PagerDuty. These are both a part of the PagerDuty app and can be added into any Datadog dashboard. Now you can quickly see which incidents are active for your team, understand how they're affecting specific business services, go ahead and investigate the root cause and finally resolve the incident, all without leaving Datadog. If you need to take more advanced actions, these links will bring you right into PagerDuty, maintaining the context of your Datadog dashboard. Now to show you how you can continue doing more with apps, I'm going to hand it over to our friends at LaunchDarkly.

Thanks, Ana. LaunchDarkly's feature flag management platform lets you manage your software in production without the risk. Today, I'm so excited to be showing off the LaunchDarkly app for Datadog. Bringing LaunchDarkly into Datadog means that you can preplan your feature releases and your associated monitoring in one place. So you can monitor while you toggle. So let's get to it. I've got my staging environment ready. I'm going to go ahead and push this change live right now, and it's done. This isn't fake. These are live changes that happen in just 200 milliseconds. So you can see the flagged change event right here as well. What's awesome now is that as I move to production, template variables are going to pull up their production flags in my widget without any extra clicks. So let's push that live, and that's it. It's live in production. Now that I can see the flagged event and the change in the API response times, that doesn't look good. So I suppose I should go ahead and roll that back. And luckily, because this is using LaunchDarkly, production changes take less than 200 milliseconds with our feature flags. So now this is completely off of production. And that's it for our quick demo. We're really happy to say that this feature is rolling out to all users today. Building an app for Datadog has been an absolute breeze, and it's really awesome to be bringing together the capabilities of our products. Back to you, Ana.

Thanks, Brandon. I'd like to give a huge shout-out to the LaunchDarkly team for being one of our pioneers for Datadog Apps and creating something that can really improve the way that our users shift their software. So with PagerDuty and LaunchDarkly, you've now seen that apps can be built out with custom dashboard widgets, side panels, models, custom context menus and even more. To conclude, let's talk a bit more about where you can find and build apps, both at Dash and beyond. For all of our users, you can find and install PagerDuty and LaunchDarkly's apps as well as others starting today in the Datadog integrations and marketplace pages. If you're a technology partner interested in building an app, be sure to attend our workshop extending the Datadog UI with apps where you'll learn how to use our brand-new developer platform. Make sure to also stop by our marketplace and apps booth in the exhibit hall to see Apps in action. Thank you.

Thanks, Ana, and thank you to all of our partners who participated in the launch of Datadog Apps. I look forward to seeing many more of you launching apps in the Datadog marketplace. Now speaking of partnerships, our own partnership with our cloud providers has been a key part to our success over the last decade. I first met our next speaker when we were just a 50-person company and his company was just starting to build out their cloud offering. Over the years, we developed a close partnership, which culminated in us starting a full Datadog region in their cloud with a tight customer experience for our DevSecOps users. Please welcome Corey Sanders from Azure.

Hi there. I am so happy to get to be a part of Dash and talk with all of you. We've come together to learn about and share some of the most exciting stuff in the tech industry, how we build and scale the next generation of technology that will empower and accelerate business and cultural innovation. It is so exciting to me for a couple of reasons. First, it's technology, and we all love technology. The opportunity to see something new, get excited about it and then be able to take it, build, adopt and adapt it to something new that I myself built is just an awesome experience and key to the growth of technology. Second, it's exciting in a much more terrifying way to me, which is that we're long past the days when IT was just the backroom department. They gave you a computer, reset your password, make sure you got your email on time. Today, every company has become a tech company. The CMO and the CIO are joined at the hip to make sure business innovation becomes synonymous with digital transformation. The companies that I get to work with, it's just so much fun to see how companies as different as retail and energy, finserv and health are all building these digital factories focused on creating new products and building out new services. It's just astounding to see the fact that software is everywhere. Now you can see this in a lot of different ways. One, of course, is that actually more than half of software engineers on the planet are working outside of the tech industry. In fact, the automotive industry hired 35% more software engineers than mechanical engineers this last year. And it just proves the urgency to enable innovation has changed how we think about it. Innovation is no longer the exception; it's the expectation. Increasingly, the expectation to deliver innovation that falls to developers and IT pros like all of us. It really is such a different industry from when I started as an intern fixing bugs in Windows. Certainly, I was different and Windows was quite different. But I've been so fortunate in my career path that I've been able to cut across a bunch of different products and services at Microsoft. And I've had sort of a front-row seat as the entire industry changes and evolves. And one of the cool things that I think that's come out of this is the increasing accessibility of technology that everyone is able to build things and do things that previously only a few could do. And it's not just all those developers who work in all these different industries that are doing such amazing work, it's also the increase in focus on local capabilities or even no-code applications that will allow any employee at a company to leverage their creativity and build an app, a solution, without having any formal training in development. It's fantastic. And while these motions are booming, both the low-code, no-code opportunities and, of course, the demand for skilled developers are going through the roof. Everyone is busy, and everyone is excited about the changes that we can make. Now while a lot of things has changed over these last 15 years of my time at Microsoft, a couple of things have remained pretty core for Microsoft. One is that, look, we are a platform company. Our focus has always been about enabling innovation for our customers. Originally, of course, it was from a box that you buy from a shelf. And now today, our platform is the Microsoft cloud, which we believe is the industry's most comprehensive and secure cloud offering. The other core constant that we've seen at Microsoft is the focus that we have with our partners. We would not be the company we are today across the decades of products and solutions without the influence and impact that our partners have had. And our relationship with Datadog is such a perfect example of why we force these partnerships. Over the last few years, we've combined complementary strengths to extend our capabilities beyond what perhaps either of us could have achieved on our own. The Microsoft cloud is a fantastic platform from edge to cloud, and our infrastructure wraps around the globe. We extend from the seafloor to space. It's a platform for creation and transformation. It is about supporting the creativity of our customers, so that they can achieve great things. It is a digital toolbox, if you will, that empowers customers to build truly amazing things. And this is where the partnership with Datadog is so amazing because Datadog sees and actually enables that potential. They see what customers are building. They see the opportunity for specialized tools to help those customers achieve what they're trying to achieve and reach that level of digital transformation. It is truly a better-together story and a story of deep mutual trust. Datadog has native integration in Azure. This hit general availability back in June and really was the first of its kind of a partner service that was available directly within the experience of the public cloud provider. It has the full seamless experience of a native Azure service, allowing you to deploy and manage and leverage the power of Datadog across all of your services and solutions just like any other Azure service. Azure is the first-and-only cloud to offer such a simplified configuration and management experience for our partners. It reduces your time as a developer to configure and manage. It makes it much, much faster to get value from the services you're deploying. And it really has enabled digital transformation at a much, much faster rate. Now for some companies, that transformation is as much a cultural change as a technology change. Organizations often have a huge amount of data, but the information they're pulling from it is poor. And so in some ways, digital transformation has really become, for many customers, a data transformation. It's not unusual to see customers with hundreds and even thousands of applications running across a diverse infrastructure. That includes a mix of owned data centers, hosting environments, branch offices, 2 or maybe even more public clouds. Such a complex story for almost every company. And to add to this complexity, customers are using different development tools, languages, frameworks as well as technologies like DevOps, Kubernetes and AI. Trying to bring them all together can be a headache. And this complexity often emerges through organic accretion of systems and apps as companies grow and acquire their own companies and solve their own problems in individual ways. Sometimes workloads can't be moved to the cloud, maybe due to regulatory reasons or data sovereignty requirements. This is very common in many of our highly regulated industries, such as financial services, health care or government. Sometimes workloads like edge workloads require ultra-low latency to the end customer, like store-based solutions for retail. This can feel like a barrier to accelerate their digital transformation and evolving the business. But in both cases, they aren't. Azure was very much purpose-built for these complex multi-cloud and hybrid environments, providing unique capabilities that give customers the flexibility to innovate anywhere in their environment and have the ability to launch multi-cloud environments, multi-locations, hybrid, on-prem and sovereign, all with one experience: saving time and cost. The amount of data created over the next 3 years will be more than all the data created over the past 30. Just to say that again, the next 3 years will be more in the past 30. We already generate data faster than many companies' capabilities to grow, ingest and put to work in any meaningful way. But we don't expect the complexity of the environments that data needs to flow through to get any simpler. We do expect demand to make it easier to manage and to keep growing. The response we're seeing to the Datadog integration shows the appetite for this. Datadog is a critical part of migrating and managing a modern data estate. Instead of siphoning off precious dev resources to monitor and aggregate workload reports, you can set it up with a single click. The Azure-Datadog integration is being used by thousands of customers across multiple industries, including travel, professional services, insurance, health care and media and communications. If you are a heavy Azure user, I recommend taking a look at the new Datadog integration. As a new customer, you can create a Datadog account directly in the Azure portal as easily as you would provision a virtual machine. You can even use your existing Azure subscription for invoicing. In just minutes, all of your Azure metrics and log data can be flowing through your Datadog environment from the very first time you log in. It's never been easier to get started using Datadog's world-class visualizations, machine learning-powered monitors and deep, powerful security tools. The need to simplify and add efficiencies is gaining urgency as organizations prioritize IT and dev resources. Datadog sees this and makes it a focus, which is why I'm so excited about our partnership and where it can go next. I know one area we see them taking advantage of is, in fact, GitHub. There is a huge overlap between GitHub and Datadog customers who largely want the same thing: to deliver and operate the best software systems faster and with more security and reliability. That's why GitHub uses Datadog internally to ensure GitHub services and APIs are always fast and available to support the largest developer community on the planet. GitHub Actions is the most powerful CI/CD service used on GitHub today, so naturally, we partnered with Datadog to integrate their new CI Visibility features with GitHub Actions and bring Datadog insights directly to developers actions pipelines. Together, we give developers the context they need to ship and operate software reliably at scale. These efficiencies and the highly collaborative GitHub environments are increasingly critical in a developer community focused on creating, innovating and building the next generation of technology. That's why we're at Dash to get a glimpse of what's around the corner and maybe learn a few things we can put to work today. The landscape for innovation has never been so wide open. The tools to create have never been better, the opportunities never so rich, and as the only cloud that helps organizations innovate anywhere with a single way to manage, govern and control multiple clouds and multiple environments, we are so excited about co-innovating on that next generation. Thank you all so much, and have a wonderful, wonderful conference.

Thanks, Corey. So many first of their kind of launches with you and your teams at Microsoft. From cloud to source control and CI, it's amazing to see our teams work as one on these initiatives. Now so far in the program, we've spent quite a bit of time talking about infrastructure, operations and developer experience. But Dash isn't just about scaling up and speeding up. It's also about security. Last year at Dash, we announced the security monitoring and CSPM offerings, our first steps into breaking down the silos between DevOps and your security teams. Since then, we've had many more developments covering everything from compliance to run time, allowing you to correlate between security and operational data. To talk about how we are breaking down this silo and bringing applications, security and infrastructure together, here is Pierre Betouin who leads Datadog Security.

Thank you, Amit. You know how much we care about performance and reliability, and we're equally excited about security. We believe that we are uniquely positioned to help engineering and security teams build more secure applications in the cloud. And the reason is production systems are becoming more and more complex. Today, a single transaction crosses an average of 35 different services. With hundreds of heterogeneous applications, APIs, microservices, web services, with thousands of developers pushing code now in production every day and thousands of internal cloud instances, it's getting up and down. We still protect modern services today, the way we protected legacy services back 20 years ago. IDS, firewalls, endpoint protections, security tools have not been designed for the cloud. But cloud technologies brought a set of new tech vectors that major sectors are actively trying to exploit. On the left, DevOps teams already have a lot of observability insights into the code and the run time. Their roles, exceptions, logs, data flows, Datadog already managed thousands and millions of services in real time. And on the security side now, dozens of different solutions are deployed usually at the network level, but none of them can provide code and contextual data that the DevOps teams need to operate. Traditional security solution like the ability to understand the deeper context of what's happening across the stack, which is becoming increasingly critical as things get more complex in the cloud. It's only by bringing and sharing high-quality and deep insights at the different layers that we can enable a better collaboration between those teams. And this is why having together observability and security capabilities on the same platform makes so much sense. That's what DevSecOps is about, and we are working and enabling that change for the millions of services we already monitor. So we learned from the observability side, and we decided to approach the problem the same way by providing a holistic platform to protect application, infrastructure and cloud environments, the exact same platform that thousands of engineers already use every day. And our users are now exactly one click away from integrating security into their stack. No need to deploy yet another agent. We are leveraging the same technologies that the ones already deployed by our customers in production. Last year at Dash Con, we announced the first product of our Cloud Security Platform. Our cloud team automatically detects critical threats for systems in production. Our users are notified whenever suspicious activities are detected in their links. For instance, when a host starts enumerating storage buckets. And unlike any traditional sim, Datadog will automatically provide the related observability data to help teams collaborate: information about EC2 instances, CPUs, data flows, users, down to the stack traces, and all of this in one click. This year, we announced the official release of our Cloud Security Posture Management product. Cloud environments are changing several times a day, putting the infrastructure at risk whenever a misconfiguration is pushed in production. Our users now get continuous release across their cloud accounts, host and containers. CSPM can, for instance, detect any change in an S3 bucket configuration that will open a new door for attackers. And we also announced the official release of our Cloud Workload Security product this year. Suspicious files or processes activities will be detected across hosts and containers and will be reported in real time with rich context. There is also one more thing that we'd like to share with you today. But for that, I'll hand over to Arnaud, Product Manager in the Application Security Group. On you, Arnaud.

Thanks, Pierre. Today, we are very excited to announce the beginning of the private beta only 6 months after the acquisition of Sqreen by Datadog. As Pierre mentioned, Application Security is now one click away with Datadog. It leverages the Datadog library, the same one powering the Gartner award-winning APM product. As it's already deployed on your production services, there is no need to allow yet another component in your stack. All it takes is setting the right environment variable and restarting your application. Out of the box, Datadog Application Security monitors hundreds of attack patterns, covering attack vectors, strengthening modern distributed system the most, such as server-side request forgery, SSRF, or injections. No manual configuration or management is required. It just works. All right. Let's see Datadog Application Security in action. Most attacks are automated scan and are nowhere close to breaching PIIs or sensitive data. They just create noise and bug down security teams. Datadog Application Security is the first security solution, leveraging distributed tracing to help teams focus on attacks that matter. In this example, Datadog Application Security has detected SSRF attempts in some distributed traces on the production service route executing network queries. In our future iteration, Datadog Application Security will go further by tracking if any user inputs actually end up in the network query, turning it into a natural SSRF vulnerability exploit. Leveraging distributed traces, teams can investigate and notify quickly the targeted business logic to evaluate the risk to the business. Datadog Application Security is part of the broader Cloud Security Platform presented by Pierre earlier. A few seconds after Application Security has detected an attempt to exploit a necessary vulnerability, Cloud Workload Security detected a share execution as route from the Azure open management infrastructure process. Combined together, application and Cloud Workload Security have actually detected an actual exploit of the OMIGOD vulnerability disclosed in September '21. Detecting such threats with confidence is the required step toward protecting your production system in an automated fashion. To recap what we just walked you through together, Datadog Application Security helps you cut through the noise by leveraging distributed traces, correcting by the unified library you're already using. It breaks silos and brings developers and security teams together on a unified platform. It's setting beyond monitoring, providing automated protections. This is an early preview of Datadog Application Security private beta. If you are interested in using this product, please go ahead and apply directly through your Datadog dashboard. Enjoy the rest of Dash Con. Thank you.

Thanks, Pierre and Arnaud. Application security, infrastructure security and compliance altogether, and all using the same agent and the same libraries that are already in your production environments. This is a perfect example of what we are looking to accomplish with Datadog. Now so far, we've shared new ways for Dev, Sec and Ops teams to break their silos and collaborate with each other over Datadog. But it wouldn't be Dash if we didn't break another silo and find new ways for your organization to benefit from observability and data. Product management, marketing and support are other parts of your organization that always want to understand how customers are engaging with applications. And to show you how we are bringing visibility and observability to these teams, I'll hand over to Miranda Kapin.

Your entire organization is designed to support the needs and happiness of your customers. Even with all the resources at your disposal, it still takes too long to understand the challenges users have with your application. And as hard as you try, this can still lead to angry customers. Say your support team is working on a high-priority case. Your customer needs to first write up their problem and then the support team needs to interpret their message, which often is missing key parts, which leads to a frustrated customer. This happens all over again when engineering product and any other team needs to get involved. You are now spending too much time gathering details from your customer, which is repetitive and frustrating. Understanding user behavior spans many teams who are troubleshooting and determining the business impact of these issues. Development process, application performance and business performance are intimately intertwined, and Datadog is committed to provide the gold standard solution to connect these steps for better organizational efficiency. This is why today, I'm proud to announce Session Replay, a great milestone in bridging these workflows. Session Replay is a video-like replay of exactly what your users are experiencing. You are now able to see and feel what happened alongside all the views, errors, logs and all the technical information that comes along with the user session in Datadog real user monitoring. Now when somebody from support receives a ticket, they can go into RUM, find that individual session, follow along with what that user is describing rather than the tedious error-prone back and forth with the customers. And it's then simple to share with the developers exactly what happened without needing to play telephone. Every person involved has the same understanding of what happened, sharing the same accurate information. The replay can be linked to a card for developers, so that if a PM wants to understand what's going on, they can quickly watch, see what happened to understand how severe the problem was. They can also highlight the lessons learned with designers by sharing a link starting at that exact point in time without them needing to rewatch the whole session. Fewer lengthy descriptions, faster problem-solving. Datadog can also help you be more proactive. In the example before, the bug could have been discovered by proactively tracking the important conversion rates in your workflow. Using funnel analysis, you identify a recent drop in usage on the individual step. You can then dive in, choose one of the sessions where the user did not successfully convert and jump right into Session Replay to investigate and learn more. Session Replay means less angry customers, faster support and less wasted engineering time. Combined with funnel analysis means both a greater understanding of how your users are reacting to your applications as well as an improved user experience. We know some of you might be worried about the data potentially being captured. I'd like to also mention that Session Replay comes with default privacy settings. It includes 3 levels of privacy: mass call, mass user input and allow all, all of which are starting points to ensure that you reliably minimize the risk of capturing sensitive or unwanted data from your end users. Session Replay is now generally available to use alongside of RUM, an exact video-like replay of your customers' experience, privacy-first options to scrub your data while controlling the security of your users' data. Funnel analysis is also now available and can be used on all of your behavioral data. Session Replay will be coming with the full RUM suite and is priced at $1.80 per 1,000 sessions. Thank you.

Thanks, Miranda. I think that's the sound of another silo being knocked down. But wait, there's one more coming. As organizations migrate from the cloud, the way we think about operational cost has changed. Now with just an API call, we scale out instantly from tens of systems to thousands across multiple clouds, achieving global scale in seconds or minutes. This changes the way finance, engineering and operations have to collaborate and makes operational data critical as they look to understand their cloud spend and how to optimize it. To share more about how we are bringing finance and operations closer together and knock down the final silo for Dash 2021, here's Jimmy Caputo.

I'm Jimmy Caputo, Director of Product Management here at Datadog. Organizations of all sizes are deploying their applications to the cloud for flexibility. With hundreds of managed services and on-demand scaling, businesses can innovate faster and respond more quickly to changing demands. But the breadth of available services and ease of adoption introduce a new challenge. Cost management is exceptionally difficult. Basic questions like how much a team spends or what a product costs are difficult to answer, especially with the move to shared compute platforms like Kubernetes. And the feedback loop between finance and engineering is broken. Finance managers and analysts can't tell if growing bills are due to increased usage or lower efficiency. Engineers are constantly making changes to their apps and infrastructure with no visibility into the cost consequences. Finally, there's widespread mistrust in opaque savings recommendations from stand-alone tools that don't know the nuances of specific workloads. To overcome these challenges, many organizations have started forming centralized FinOps teams whose sole responsibility is breaking down silos and bringing discipline to cloud cost management. But for these FinOps teams and their peers in finance and engineering to be successful, everyone needs access to timely and granular cost and operational data in a single pane of glass. That's why we're excited to announce the beta launch of Datadog Cloud Cost Management. Datadog Cloud Cost Management gives FinOps teams, finance and engineering unprecedented visibility into cloud costs. With just a few clicks, Datadog starts ingesting your cloud cost data, transforms it into easy-to-query metrics and enriches it with the observability data you're already sending to the platform. You get actionable insight into spend by team, product or any other dimension you care about for both dedicated and shared resources. When costs rise, you can quickly correlate the change with usage metrics to determine the root cause. And with timely cost and observability data visible together, engineering teams know the financial impact of their changes and can implement actions to save with confidence. Let's take a look at a quick demo. We're looking at the new cloud costs UI in Datadog, which lets cost managers quickly filter in group cost data to hone in on what's changing over time. On top, we see the current query and just below that, a summary graph and table with the results. On the left, we have a set of predefined facets that we can use to quickly drill down. Let's take a look at the amortized cost grouped by cloud product over the last 4 months. Looking at the table, I can see that AWS Lambda has seen the biggest increase. I'm going to drill into that to understand a bit more about what's going on. In the side panel, I get a more detailed look at my Lambda costs. Let's take a look at the weekly rollout to see precisely when things changed. When I clicked on the Metrics tab, the Datadog metrics for my Lambda functions are displayed for the same time window. It looks like the number of executions hasn't increased, but the duration of my functions has been steadily rising. This is probably what's driving the increased cost. I can further group by tags like team and function to continue investigating the spike. From here, I can also export my analysis to a dashboard or download a screenshot to share with the team and determine if this was intentional or an error. In addition to the new UI, you can add cloud costs to widgets in your existing Datadog dashboards. As an engineering manager, this lets me track the financial and operational health of my team's services in a single view. In this dashboard for an e-commerce checkout service, I can see KPIs that are important to my business, check out volume and success rate. Right next to it, I can see the total cost to run the service, the cost by cloud product and the cost per transaction processed, all calculated from metrics in Datadog. Further down, I can see even more details about my infrastructure spend and related utilization metrics, which helps me spot potential inefficiencies and savings opportunities. With Datadog Cloud Cost Management, FinOps teams, finance and engineering have a new level of visibility into their cloud costs, enabling them to understand trends, allocate spend across the organization and identify inefficiencies. If you're interested, Cloud Cost Management is available as a private beta. You can visit the URL shown here to request access. Thank you.

Thanks, Jimmy. We're excited to welcome finance and FinOps silos into the fault. Now we've covered a lot of ground today, including tools that let you ingest and store high-volume data sets without giving up control or freedom of where your data lives; new rich applications that let our partners build new workflows directly from dashboards; CI Visibility and source control integrations that bring observability data closer to your development workflows; and application security and observability together to help you protect your most critical systems. And of course, we found new silos to tear down and friends to invite to our observability journey. But that's just the start of Dash. Over the next few days, you'll see more releases and capabilities launch that help you build your observability into every part of your organization. You'll also have opportunities to network with and learn from fellow Datadog community members and get hands-on with all our launches during our workshops or check out live demos in the virtual exhibit hall. We look forward to learning with you as you scale up, speed up and secure your applications at Dash. Thank you.