Good day, and thank you for standing by. Welcome to the MongoDB Third Quarter Fiscal Year 2025 Conference Call. At this time, all participants are in a listen-only mode. After the speakers' presentation, there'll be a question-and-answer session. Please be advised that today’s conference is being recorded. I would now like to turn the call over to your speaker for today, Brian Denyeau. Please go ahead.

Thank you, Lisa. Good afternoon, and thank you all for joining us today to review MongoDB's third quarter fiscal 2025 financial results, which we announced in our press release issued after the close of the market today. Joining me on the call today are Dev Ittycheria, President and CEO of MongoDB; and Michael Gordon, MongoDB's COO and CFO. During this call, we will make forward-looking statements, including statements related to our market and future growth opportunities, our expectations for the macroeconomic environment in fiscal 2025, and the impact of AI, the benefits of our product platform, our competitive landscape, customer behaviors, our financial guidance, and our planned investments in growth opportunities in AI. These statements are subject to a variety of risks and uncertainties, including the results of operations and financial conditions that could cause actual results to differ materially from our expectations. For a discussion of the material risks and uncertainties that could affect our actual results, please refer to the risks described in our quarterly report on Form 10-Q for the quarter ended July 31st, 2024, that we filed with the SEC on August 30th, 2024. Any forward-looking statements made on this call reflect our views only as of today, and we undertake no obligation to update them except as required by law. Additionally, we will discuss non-GAAP financial measures on this conference call. Please refer to the tables in our earnings release in the Investor Relations portion of our website for a reconciliation of these measures to the most directly comparable GAAP financial measure. With that, I'd like to turn the call over to Dev. Dev?

Thanks, Brian, and thank you to everyone for joining us today. I'm pleased to report that we had a strong quarter of new business and executed well against our large market opportunity. Let's begin by reviewing our third quarter results before giving you a broader company update. We generated revenue of $529 million, a 22% year-over-year increase and above the high end of our guidance. Atlas revenue grew 26% year-over-year, representing 68% of total revenue. We generated non-GAAP operating income of $101 million for a 19% non-GAAP operating margin, and we ended the quarter with over 52,600 customers. Overall, we were pleased with our performance in the third quarter. We had a strong new business quarter and we're happy with our new workload acquisition on Atlas. Our non-Atlas business significantly exceeded expectations in part because we benefited from a few large multi-year deals as customers continue to value our run anywhere strategy and want to build a deeper, longer-term relationship with MongoDB. Atlas consumption was slightly better than expected in a macro-environment that we would characterize as largely consistent with what we saw in the first half of the year. Michael will cover consumption trends in more detail. Retention rates remained strong in Q3, demonstrating the mission criticality of our platform. On our Q1 earnings call, we shared with you the three major strategic initiatives that we believe will enable us to maximize our long-term opportunity. I want to give you an update on the progress we're making on those initiatives. First, we are increasing our investment in the enterprise channel since we see the strongest returns in this part of the market. Specifically, we're expanding our strategic account program going into next year, as we see more accounts that will benefit from incremental investment. In addition, we're investing time and resources to educate developers in large enterprise accounts and uplevel their MongoDB skills. These organizations have thousands of developers, and as we penetrate them more deeply, we encounter developers who have historically only built SQL applications and simply do not know how to use MongoDB to its full potential. In our experience, educating these developers on the benefits of MongoDB drives significant incremental adoption of our platform. To fund these up-market investments, we are reallocating a portion of our mid-market investments. The mid-market remains an attractive opportunity for us, but we believe that prioritizing investment up-market will deliver strong returns in the current environment. We also believe there are additional ways to serve the mid-market more efficiently through our self-serve channel and other scaled technology-enabled sales and customer service motions. Second, we are optimistic about the opportunity to accelerate legacy app modernization using AI and are investing more in this area. As you recall, we ran a few successful pilots earlier this year, demonstrating that AI tooling combined with professional services and our relational migrator product can significantly reduce the time, cost, and risk of migrating legacy applications onto MongoDB. While it's early days, we have observed a more than 50% reduction in the cost to modernize. On the back of these strong early results, additional customer interest is exceeding our expectations. Large enterprises in every industry and geography are experiencing acute pain from their legacy infrastructure and are eager for more agile, performant, and cost-effective solutions. Not only are customers excited to engage with us, they also want to focus on some of the most important applications in their enterprise, further demonstrating the level of interest and size of the long-term opportunity. As relational applications encompass a wide variety of database types, programming languages, versions, and other customer-specific variables, we expect modernization projects to continue to include meaningful service engagements in the short and medium term. Consequently, we're increasing our professional services delivery capabilities, both directly and through partners. In the long run, we expect to automate and simplify large parts of the modernization process. To that end, we are leveraging the learnings from early service engagements to develop new tools to accelerate future modernization efforts. Although it's early days and scaling our legacy app modernization capabilities will take time, we have increased conviction that this motion will significantly add to our growth in the long term. Third, we are investing to capitalize on our inherent technical advantages as a key component of the emerging AI tech stack. As a reminder, MongoDB is uniquely equipped to query rich and complex data structures typical of AI applications. The ability of a database to query rich and complex data structures is crucial because AI applications often rely on highly detailed, interrelated, and nuanced data to make accurate predictions and decisions. For example, a recommendation system doesn't just analyze a single customer's purchase but also considers their browsing history, peer group behavior, and product categories requiring a database that can query and interlink these complex data structures. In addition, MongoDB's architecture unifies source data, metadata, operational data, and vector data in an all-in-one platform, outdating the need for multiple database systems and complex back-end architectures. This enables a more compelling developer experience than any other alternative. From what we see in the AI market today, most customers are still in the experimental stage as they work to understand the effectiveness of the underlying tech stack and build early proof-of-concept applications. However, we are seeing an increasing number of AI apps in production. Today, we have thousands of AI apps on our platform. What we don't yet see is many of these apps actually achieving meaningful product market fit and therefore significant traction. In fact, as you take a step back and look at the entire universe of AI apps, a very small percentage of them have achieved the type of scale that we commonly see with enterprise-specific applications. We do have some AI apps that are growing quickly, including one that is already a seven-figure workload that has grown 10x since the beginning of the year. Similar to prior platform shifts, as the usefulness of AI tech improves and becomes more cost-effective, we will see the emergence of many more AI apps that do nail product market fit, but it's difficult to predict when that will happen more broadly. We remain confident that we will capture our fair share of these successful AI applications as we see our platform is popular with developers building more sophisticated AI use cases. We continue investing in our product capabilities, including enterprise-grade Atlas Vector Search functionality to build on this momentum and even better position MongoDB to capture the AI opportunity. In addition, as previously announced, we are bringing search and vector service to our community and EA offerings, leveraging our run anywhere competitive advantage in the world of AI. Finally, we are expanding our MongoDB AI applications program, or MAAP, which helps enterprise customers build and bring AI applications into production by providing them with reference architectures, integrations with leading tech providers, and coordinated services and support. Last week, we announced a new cohort of partners including McKinsey, Confluent, Capgemini, and Unstructured, as well as the collaboration with Meta to enable developers to build AI-enriched applications on MongoDB using Llama. Next, I'd like to provide you with a brief product update. At our dot local developer conference in London in October, we announced the general availability of MongoDB 8.0, the fastest and most performant version of MongoDB ever. MongoDB 8.0 performs 20% to 60% better against common industry benchmarks compared to our prior version and is built to exceed our customers' most stringent security, resiliency, availability, and performance requirements. To best serve our customers, we regularly review and reprioritize investments in our product portfolio to ensure we're allocating our resources to products with the highest demand from our customers. And to do that, we also deprecate products that are not showing results we desired. Consequently, we made the decision to consolidate our Atlas serverless offerings with our smallest dedicated tiers to create Atlas Flex customers, a new offering with a simpler architecture that provides the elasticity features akin to serverless. We will begin migrating effective customers to the single, simple entry-level solution in Q4. We also decided to deprecate Atlas Device Sync and other capabilities not widely adopted in order to focus our engineering resources on the core platform. While these reprioritization decisions are not made lightly, they allow us to deliver the most value to the largest number of customers, reinforcing our commitment to being the best modern database and helping us to grow faster. Now, I'd like to spend a few minutes reviewing the adoption trends of MongoDB across our customer base. Customers across industries and around the world are running mission-critical projects in MongoDB Atlas, leveraging the full power of our developer data platform, including Financial Times, CarGurus, and Victoria's Secret. As part of the digital transformation journey, global specialty retailer Victoria's Secret & Company migrated its e-commerce platform to MongoDB Atlas. As a fully managed platform, MongoDB Atlas allowed the company to simplify its architecture and improve performance, supporting the retailer to provide a resilient, secure, and fast web and mobile e-commerce experience for their millions of customers around the world. Allianz, Alfamart, Swiss Post, and Paylocity are turning to MongoDB to modernize applications. Paylocity, a leading provider of cloud-based payroll and human capital management software, selected MongoDB to power proprietary application aimed at fostering employee connections and engagement. When traffic increased and the original SQL-based solution was unable to keep up with the required performance metrics, Paylocity migrated to MongoDB Atlas to take advantage of the flexible schema architecture, performance, and scalability. MongoDB costs five times less than the previous SQL database solution, and the company's developers can now create an application within minutes, something that used to take weeks. Mature companies and startups alike are using MongoDB to help deliver the next wave of AI-powered application to customers, including NerdWallet, Cisco, and Tealbook. Tealbook, a supplier intelligence platform, migrated from Postgres, PG Vector, and Elastic Search to MongoDB to eliminate technical debt and consolidate their tech stack. The company experienced workload isolation and scalability issues in PG Vector and were concerned with the search index inconsistencies, which were all resolved with the migration to MongoDB. With Atlas Vector Search and dedicated search nodes, Tealbook has realized improved cost-efficiency and increased scalability for the supplier data platform, an application that uses GenAI to collect, verify, and enrich supplier data across various sources. In summary, we had a healthy Q3 with both Atlas and EA exceeding expectations. We saw a strong new business quarter and we remain confident in our ability to become an increasing strategic provider in our large and growing market. Looking forward, we see a great opportunity to grow our adoption in the enterprise through new workloads, modernizing legacy applications, and winning the next generation of AI-powered applications. I would like to finish by providing an update on our senior leadership. First, as we announced early in the press release, after nearly 10 years, Michael Gordon has made the decision to leave MongoDB. Michael has been instrumental in MongoDB's success over the past decade, leading our successful IPO, helping us grow our revenue nearly 50-fold, and successfully scaling our business model to generate meaningful operating leverage. He has been a trusted advisor and business partner to the Board and me over the years and also has become a personal friend. Michael is excited to take a well-deserved break. We have commenced the search for Michael's replacement and will be evaluating both internal and external candidates. One of Michael's proudest accomplishments has been building a world-class finance team under his leadership, and I'm confident that we will not miss a beat during this transition. Michael will continue to serve as CFO through January 31st to help us finish the fiscal year and then will transition to an Advisor to the company to ensure a seamless process. If we have not named Michael's successor by fiscal year-end, Serge Tanjga, SVP of Finance, will serve as Interim CFO, beginning on February 1st. Second, we are promoting Cedric Pech, currently our Chief Revenue Officer, to the newly created role of President Worldwide Field Operations. In this new position, Cedric will oversee all our field-based customer-facing and go-to-market enablement teams, including professional services. We believe this org structure will best enable us to execute on some of the key strategic initiatives I discussed earlier, in particular, our increased focus on up-market and the app monetization opportunity. I would like to congratulate Cedric on this well-deserved promotion. With that, let me turn the call over to Michael.

Thanks, Dev, and thanks for the kind words and our incredible partnership over the past decade. The past 10 years have been the most rewarding in my professional career, and I'm extremely proud of what we've achieved together and of course, to the whole MongoDB team. With as much success as we had, I still believe that MongoDB is in the early stages of realizing its full potential as it continues to take share in one of the largest markets in software. Now, turning to the results for the quarter. I'll begin with a detailed review of our third-quarter results and then finish with our outlook for the fourth quarter and full fiscal year 2025. First, I'll start with our third quarter results. Total revenue in the quarter was $529.4 million, up 22% year-over-year and above the high end of our guidance. Shifting to our product mix, Atlas grew 26% in the quarter compared to the previous year and now represents 68% of total revenue compared to 66% in the third quarter of fiscal 2024 and 71% last quarter. We recognized Atlas revenue primarily based on customer consumption of our platform and that consumption is closely tied to end-user activity of their applications. Let me provide some context on Atlas consumption in the quarter. In Q3, consumption was slightly ahead of our expectations. This year's Q3 seasonal improvement was more muted than in years past as expected. On a year-over-year basis, consumption growth remains below that of the prior year period. Turning to non-Atlas revenue. Non-Atlas came in significantly ahead of our expectations. As Dev mentioned, EA new business was strong, and we continue to have success selling incremental workloads into our existing customer base. In addition, our Q3 non-Atlas revenue benefited from a few large multi-year deals. As you know, due to ASC 606, we recognized the entire term license component of a multi-year contract at the start of that contract. Compared to Q3 of last year, the multi-year license component of non-Atlas revenues was over $15 million higher. Turning to customer growth. During the third quarter, we grew our customer base by approximately 1,900 customers sequentially, bringing our total customer count to over 52,600, which is up from over 46,400 in the year-ago period. Of our total customer count, over 7,400 are direct sales customers, which compares to over 6,900 in the year-ago period. The growth in our total customer count is being driven primarily by Atlas, which had over 51,100 customers at the end of the quarter compared to over 44,900 in the year-ago period. It is important to keep in mind that the growth in our Atlas customer count reflects new customers to MongoDB in addition to existing EA customers adding their first Atlas workload. Continuing on, in Q3, our net ARR expansion rate was approximately 120%. We ended the quarter with 2,314 customers with at least $100,000 in ARR and annualized MRR, up from 1,972 in the year-ago period. Moving down the income statement, I'll be discussing our results on a non-GAAP basis unless otherwise noted. Gross profit in the third quarter was $405.7 million, representing a gross margin of 77%, which is flat versus the year-ago period. Our income from operations was $101.5 million or a 19% operating margin for the third quarter compared to an 18% operating margin in the year-ago period. The primary reason for more favorable operating income results versus guidance is our revenue outperformance, including the very high-margin multi-year license revenue benefit. Net income in the third quarter was $98.1 million or $1.16 per share, based on 84.2 million diluted weighted average shares outstanding. This compares to a net income of $79.1 million or $0.96 per share on 83.7 million diluted weighted-average shares outstanding in the year-ago period. Turning to the balance sheet and cash flow, we ended the third quarter with $2.3 billion in cash, cash equivalents, short-term investments, and restricted cash. Operating cash flow in the third quarter was $37.4 million. After taking into consideration approximately $2.9 million in capital expenditures and principal repayments of finance lease liabilities, free cash flow was $34.6 million in the quarter. This compares to free cash flow of $35 million in the year-ago period. In Q3, we did not incur capital expenditures to purchase IPV4 addresses as we previously expected, but we did start making those purchases in November and still expect a total outlay of $20 million to $25 million this fiscal year as we previously communicated. I'd now like to turn to our outlook for the fourth quarter and full fiscal year 2025. For the fourth quarter, we expect revenue to be in the range of $515 million to $519 million. We expect non-GAAP income from operations to be in the range of $55 million to $58 million and non-GAAP net income per share to be in the range of $0.62 to $0.65 based on 84.9 million estimated diluted weighted average shares outstanding. For the full fiscal year 2025, we expect revenue to be in the range of $1.973 billion to $1.977 billion, non-GAAP income from operations to be in the range of $242 million to $245 million, and non-GAAP net income per share to be in the range of $3.01 to $3.03 based on 84 million estimated diluted weighted average shares outstanding. Note that the non-GAAP net income per share guidance for the fourth quarter and full fiscal year 2025 includes a non-GAAP tax provision of approximately 20%. I'll now provide some more context around our updated guidance. First, in terms of Atlas consumption, we expect to see a typical seasonal slowdown in Q4, driven by underlying application usage moderating during the holiday season. Second, since Atlas consumption remained lower on a year-over-year basis in Q3, we expect to see continued deceleration of Atlas year-over-year growth in Q4. Third, we expect to see a sequential decline in non-Atlas revenue in Q4, which is contrary to our normal pattern. The reason for this is that we experienced a significant additional benefit from multi-year deals in Q3, which we do not expect to recur in Q4. In addition, I want to provide some incremental color on some of our recent product and go-to-market changes that will impact the growth of our reported customer count going forward. First, as Dev explained, we are reallocating a portion of our go-to-market resources from the mid-market to the enterprise channel. As a result, we expect to see significantly fewer mid-market direct sales customer net additions and as a result, slower direct sales customer growth going forward. We believe this reallocation of investment dollars will drive higher revenue growth over time. So, it's a trade-off that makes sense. Second, as we introduce Atlas Flex clusters in Q4 and automatically migrate customers in Q1, we expect to see a one-time negative impact to our customer count since we have approximately 4,000 serverless customers who are very low spending, and we do not expect them to transition over to Flex. These customers have a negligible impact on our revenue but will impact our reported customer count. To summarize, we're pleased with our third quarter results and especially our ability to win new business. We have a small share in one of the largest and fastest-growing markets in all of software with a number of secular tailwinds, including AI at our back. We'll continue investing judiciously and focusing on our execution to capture this long-term opportunity. With that, we'd like to open it up to questions. Operator?

Our first question for the day will be coming from Sanjit Singh of Morgan Stanley. Your line is open.

Thank you for taking the questions and congratulations, Michael, on your remarkable career. You had a fantastic tenure at MongoDB. I'm eager to see what you do next or even if you choose to take a break. So congrats, Michael. To start off with the questions, when we examine Atlas's performance over the last two quarters, please correct me if I'm mistaken, but it seems that consumption is coming in at least modestly ahead of your expectations. Compared to the start of the year, what factors are contributing to this improvement in Atlas consumption over the past two quarters? Is it related to sales execution or an increase in end-user activity?

We initially projected stable growth for Atlas consumption at the beginning of the year. However, we've observed lower year-over-year growth in actual consumption, which we factored into our Q4 guidance. While Q2 and Q3 performance exceeded our expectations, we still see a decline compared to the previous year. I want to clarify that there's a difference between year-over-year comparisons and our expectations.

Understood. And then, Dev, I have to ask you the AI agent question. In terms of an AI agent needing more context, it has going to have a set of tools to take its actions. What does it mean for MongoDB as an operational data store as customers start to roll out more agentic applications?

Certainly. Regarding agents, when we talk about them, we can consider jobs, projects, and tasks. Currently, the agents being introduced primarily focus on tasks, similar to offerings from companies like Sierra. However, their main purpose is to create rich and complex data structures. This is crucial for AI because models rely on understanding relationships, hierarchies, and patterns within data rather than just isolated data points. They need to provide real-time insights. For instance, if a customer is inquiring about an order placed just five minutes ago without receiving confirmation, the chatbot must handle real-time information effectively. Advanced use cases, such as fraud detection and understanding supply chain behaviors, require an in-depth comprehension of intricate data relationships. This aligns perfectly with what MongoDB provides, positioning us favorably in this space. Additionally, we have seamlessly integrated search and vector search capabilities, making us more than just an OLTP database. This combination creates a unique experience that no other platform can match, giving us a significant advantage. We are also compatible with leading AI frameworks and platforms, offering enterprise-grade security and compliance. Our customers can operate us in any of the 118 cloud regions or on-premises, which is another important differentiator for us.

Awesome. Appreciate the thoughts, Dev.

Thank you. One moment for the next question. And our next question will be coming from the line of Tyler Radke of Citi. Your line is open.

Hi, thank you very much for taking the question. And, Michael, all the best, and congratulations on 10 years. Going back to the sales execution, I mean, one of the things that you talked about earlier this year was some challenges just in terms of the recently-acquired workloads ramping. And I think a lot of those were from the past fiscal year. So curious how the quality of workload acquisition has trended this year. And as you think about the ramp in consumption potential into next year, how does that sort of look versus this time a year ago?

Certainly. I will discuss the changes we made and then Michael will share insights about consumption trends. We implemented some adjustments at the start of the year, focusing on both the volume and quality of the workloads. We believe these changes are having a positive impact. It's still too early to confirm success, as these workloads typically begin small and expand over time, but we are encouraged by the results we've observed so far. However, it's still early in the process. We will gain more insight into the fiscal '25 workloads as we progress toward fiscal '26. Overall, things are looking good. Michael, can you provide an update on consumption?

Yeah, just a couple of things, and thanks, Tyler. The fiscal '24 cohorts that we called out earlier, that slower growth does continue. They've been in line with our revised expectations. We made some changes that we talked about earlier in the year that should affect the fiscal '25 cohorts, but it's just too early to tell. On those, we need a few more quarters of data before you can really see how we're seeing those behave differently. I will say we've talked about the new business environment and our success in new business. We have been pleased with that, but that just shows kind of the initial piece and we need to see how they grow and how those cohorts evolve.

Thanks. Regarding the EA side, you mentioned the significant strength in the non-Atlas business this quarter. Could you elaborate on the relative upside attributed to duration as opposed to new business? I know you mentioned the year-over-year impact of duration, but do you see this as a one-time occurrence, or do you think that some of your larger customers are increasingly leaning towards EA? How does this influence your perspective on the product and the introduction of features like vector search and stream processing to the on-prem product?

Yeah, thank you. So overall, we continue to find the EA product resonate with customers. It's an important part of the run anywhere strategy, and we've continued to see success with that and people wanting to increase their investment in MongoDB. There's always been a multiyear component and we continue to see that. We talked about that at the beginning of the year as to how fiscal '24 had an abnormally high amount of multi-year benefit and therefore, we were anticipating that being a headwind and we quantified that in roughly the $40 million range. What we talked about on this call earlier is we saw from a few large accounts, a surprising amount of multiyear that positively benefited Q3 at a little more than $15 million in revenue compared to what we saw in Q3 a year ago. So, not as much of a headwind as we had been expecting. Obviously, with the 606 dynamics, some of these things, especially for a large deal can be kind of meaty and chunky and lumpy, which is why we try and call it out and sort of help people understand, but there's a pretty healthy kind of baseline flow, not just of EA but also of multiyear. And when we see spikes, we just try and call it out for you.

Yeah. I just want to add, Tyler, that we are investing in what we call our EA business. First, we're starting by investing with search and vector search and a community product. That does a couple of things for us. One, whenever anyone starts with MongoDB with the open-source product, they immediately get all the benefits of that complete, highly integrated platform. Two, those capabilities will then migrate to EA. So, EA for us is an investment strategy. We definitely see lots of large customers who are very, very committed to running workloads on-prem. We even see some customers want to run AI workloads on-prem. So, the optionality they get by using MongoDB to not just be on-prem and the cloud, but also cross-cloud, is a very compelling one.

Thank you.

Thank you. One moment for the next question. And our next question will be coming from the line of Brad Reback of Stifel. Your line is open.

Great. Thanks very much. And, Michael, best of luck. Dev, you started the call talking about a bunch of investments, which are great given the growth of the business. And obviously, you talked about reallocating some expenses. But net-net, should we think about this incremental investment phase next year as gating margin upside?

I think that's something we are not ready to discuss for next year at this moment, but I would say that we are looking to invest in the modernization of legacy apps where we see large workloads potentially available and being the ideal database for GenAI applications, which represents the future. These are significant investments that will drive long-term growth, and we feel very positive about them.

That's great. And then on the MAAP program, are most of those workloads going to wind up in Atlas or will that be a healthy combination of EA and Atlas?

I think it's still early days. I would lean more towards Atlas and EA in the beginning. Once we add search and vector search capabilities to the EA product, you will see more on-prem use. Clearly, people can utilize MongoDB for AI workloads with other technologies alongside MongoDB for on-prem AI applications. However, I believe that you will likely see this trend develop first in Atlas.

Great. Thanks very much.

Thank you.

Thank you. And one moment for the next question. Our next question will be coming from the line of Jason Ader of William Blair. Your line is open.

Yeah, thank you. I'm not going to belabor congratulating Michael, but it has been fun working with you and best of luck. The question I had is on the strength in EA. Do you think, Dev, it represents a comment on how enterprises might be rethinking or reassessing the kind of on-prem versus cloud workload placement decision?

Well, when I think about large enterprises, I think large enterprises have meaningful workloads that are still running on-prem. I think the belief that everything would go to the cloud was probably something that was really popular in the good old days. But I think now as customers assess their investments that they already have in place, they're being much more judicious about where they run those workloads and if they think they can leverage their existing investments in their own infrastructure, then they're going to do so. Also for a bunch of other reasons like regulatory reasons, some customers are not moving as aggressively to the cloud. We see that particularly in Europe, where many of the European banks are still running the majority of their workloads on-prem. So it also varies by region where, conversely in Asia, we're seeing people move much more aggressively to the cloud. So I think it really depends on industry, on geography, and on the personal dynamics of what's happening in that particular account. I mean, we see some large U.S. banks are also very committed to running things on-prem. So it really varies. And that's why we feel really good about our run anywhere strategy because it gives customer optionality. They can build something and run on-prem, and if and when they choose to move to the cloud, it's very easy to do so with MongoDB.

All right. And then just as a follow-up also on the investments you're making in strategic sales and enterprise. Could you just get a little more specific on what those investments might be? Is it hiring a lot of new salespeople? Is it working more with systems integrators, investing more in SIs? Any additional detail would be helpful. Thanks.

Yeah. So just for everyone's benefit, we've identified a number of accounts, which we call strategic accounts, which we think that have high upside for us. We've seen a number of accounts that grow very quickly when we deploy the right mix of resources. Now they're all not necessarily quota-carrying resources. They could be additional technical sales resources, additional professional services resources, additional customer service resources to better service and support those accounts. We even do things like run education sessions for developers of the accounts. They’re called hackathons or developer days or even design reviews where we'll meet with our development teams who are looking to build an application and help them think about how they would potentially use MongoDB to build that particular app. What we find is that because many of these developers, their experience with MongoDB is quite limited, the more we can engage with them, the more we can educate them and the more we can show them how simple and easy it is. Like, for example, most customers today think like they have to use an OLTP database, then a search database, maybe a vector database, and then a caching database. And all that is integrated in MongoDB. So all of a sudden, customers can say, wow, I can simplify my life, simplify my back-end infrastructure, build this app far more quickly and it will be much more manageable long-term if I do everything on MongoDB. And it's really a function of just educating them on the power of MongoDB that really opens up a lot of opportunities for us. So that's why we're doubling down, and the mix of resources is really predicated on the accounts, but it's not just quota-carrying resources, it's the whole suite of resources that we're bringing to the table.

Thank you.

Thank you. One moment for the next question. Our next question will be coming from the line of Andrew Nowinski of Wells Fargo. Your line is open.

Okay. Good afternoon. Thank you very much for taking the question, and congrats on a nice quarter. You gave an example of a customer that migrated off Postgres, and I think you said they had issues with their PG vector function. I was wondering how long was that customer using Postgres before they decided to make a change to Mongo, meaning was this some sort of like a rebound type customer where they chose Postgres and it didn't work? And then how frequently are you seeing this type of transition?

I can't provide the exact details on how long they were using Postgres, but this is a trend we're observing in our business. It's important to note that Postgres is a 40-year-old technology, and it has benefitted from users moving away from other relational databases like Oracle, SQL Server, and MySQL. Being an open-source database, Postgres also faces the same inherent challenges as all relational databases, particularly in terms of inflexibility. Once a schema is built, altering it becomes quite difficult, and scaling and distributing data can also be complex. When dealing with large volumes of data, users often resort to unconventional solutions, like off-road storage for sizable data objects, leading to performance issues. Many people default to using Postgres simply because they lack knowledge of better options, as they are accustomed to relational databases, and there's a broader shift away from those traditional relational platforms. Once we educate developers about the flexibility of our schema, the ease of horizontal scaling, the advantages of our rich query language for aggregations, and the efficiency of the document model for data organization, many are pleasantly surprised by how much easier it is. I want to emphasize that this is not a zero-sum game; Postgres does not need to fail for us to succeed. The market is large, and we are excited about the opportunities. However, we definitely see customers transitioning from Postgres to MongoDB.

Thank you. That was very helpful. And maybe just a quick follow-up. If we normalize the $15 million multiyear deal impact you had in Q3, would EA still be down sequentially in Q4? Thank you.

We haven't given that level of guidance, but just trying to help you understand in the context of the full-year numbers and the headwind that we talked about at the beginning of the year, just given the strength that we saw in Q3.

Thank you. And one moment for the next question. The next question will be coming from the line of Raimo Lenschow of Barclays. Your line is open.

Thank you. Michael, could you share your insights on the EA performance this quarter? How should we view the renewal situation as we move from Q3 to Q4, and what implications does this have for upsell and cross-sell opportunities as clients consider initiating AI projects through self-serve, as you mentioned earlier?

I believe that Q4 is typically a significant renewal period for EA. However, our guidance indicates that due to strong performance in multiyear, we expect EA to decrease sequentially, which is unusual for us, and that’s why we highlighted it. When it comes to AI workloads and related developments, it’s too soon to make a definitive assessment. We'll continue to refine our perspective as we approach the full-year guide in March, at which point we’ll also provide insights on cohort behavior and the impact of multiyear. I hope my earlier comments regarding Q4 are helpful.

Yeah. Okay, perfect. And then can you talk a little bit about like obviously, there's a debate of like which database will be the persistent layer if you do AI projects, et cetera. What do you see from the big hyperscalers in terms of working with you guys and partnerships? We obviously just have AWS kind of summit, et cetera. Can you speak a little bit like how your relationship with those big guys is evolving around this? And Michael, all the best in case I don’t talk to you.

I'll start with the partnerships, beginning with AWS. They just had their re:Invent show last week, and our relationship remains very strong. We've closed a significant number of deals this past quarter, including some large ones. We're integrating new products like Q and Bedrock, and field engagement has been excellent. Regarding Azure, we had a slow start, but now the Azure MongoDB relationship is stronger than ever. We've secured many deals and are involved in the Azure native IC service program, with deep integrations that include Fabric, Power BI, Visual Studio, Semantic Kernel, and Azure OpenAI Studio. We are also one of Azure's largest marketplace partners. With GCP, we've seen an increase in co-sales this past quarter, and they've made some favorable changes to collaborate with MongoDB, leading to positive results. Going forward, we are focused on closing several large deals with GCP in Q4. Overall, things are going well. As for your question about hyperscalers potentially bundling their offerings with AI projects, they've been bundling their database services with all of their offerings from the beginning. This has been their main strategy, and we've executed well since databases are crucial decisions for customers. The hyperscalers recognize our performance and see the value in partnering with us. Customers understand the significance of the data layer, especially for AI applications, which reinforces our strong partnerships with all three hyperscalers.

Okay, perfect. Thank you.

Thanks, Raimo.

Thank you. And one moment for the next question. Our next question will be coming from the line of Brad Sills of Bank of America. Your line is open.

Great. Thank you so much, and congratulations, Michael, on your next move. I wanted to ask about new workloads here on vector search, stream processing, relational migrator. Is there any one of those three that's ramping faster than maybe you expected? Just a little bit of color on how those new workload types are ramping. Thank you.

Sure, let me provide you with an overview of our new products. Recently, we launched a feature called Atlas search nodes that allows users to scale their search nodes asymmetrically. This is particularly beneficial for search-intensive applications, as it reduces the cost of scaling all nodes. The response to this capability has been overwhelmingly positive, with high demand because customers can customize configurations to meet their specific search needs. For example, one of the largest banks is utilizing Atlas Search to deliver a Google-like search experience for payments data aimed at large corporate clients, making performance and scalability essential. Additionally, a prominent provider of AI-powered accounting software is leveraging Atlas Search for its invoice analytics feature, enabling finance teams and end users to conduct ad-hoc analyses and identify overdue invoices or those with errors. On the vector search front, it has been our first full year since its general availability, and the uptake has been impressive. In Q3, we introduced quantization for Atlas Vector Search, which can reduce memory requirements by up to 96%, allowing us to handle larger vector workloads with significantly better pricing and performance. A multinational news organization has developed a GenAI powered tool that allows producers and journalists to efficiently search, summarize, and verify information from extensive data sources. Moreover, a leading security firm employs Atlas Vector Search to combat AI fraud, while a major global media company has replaced Elasticsearch with our hybrid and vector search for its user recommendation engine, which suggests articles to users. We are also witnessing strong interest in our streaming product, with high demand. We recently introduced it to another hyperscaler, and customers have expressed that integrating stream processing with MongoDB significantly simplifies their operations. Overall, we are very satisfied with the advancements we are making with our new products, and as mentioned earlier, bundling these capabilities helps customers avoid the complexities and costs of integrating multiple technologies, ultimately saving them time, money, and reducing risk.

That's really exciting. Thanks, Dev. And then I wanted to ask a question around Cedric's appointment. Any focus that may be different here under his leadership that we should be thinking about going forward? Thank you.

No, Cedric has been our CRO for, gosh, now like I think five or six years, and I was the Interim CRO for about three quarters until he took over when we last made a change. This is really an expansion of his responsibilities. I've known Cedric for a long time. He and I have worked with at multiple different companies. I think I have a good barometer for understanding sales leadership. There are a number of sales leaders who worked at other top-tier software companies who used to work for me or with me. And so I'm super-excited by the role Cedric is going to take, and then we're also making some changes under Cedric to better align the different organizations so that we can more tightly work together on going up-market on app monetization and positioning ourselves well to be the ideal database for GenAI apps.

Super exciting. Thanks, Dev.

Thank you.

Thanks, Dev.

Thank you. And one moment for the next question, please. Our next question will be coming from the line of Mike Cikos of Needham and Company. Your line is open.

Hey guys, thanks for taking the question here. I just wanted to come back to the consumption growth being slightly better than expectations again for the second quarter in a row now. And apologies if I missed it, but this improvement that we're seeing, is this across all vintages and geographies or is it potentially more concentrated in scope? Just trying to get a better understanding of what's taking place out there and what's embedded in the guide.

Yeah. No, I would describe it as broad-based, Mike. And obviously, we're pleased to see it, and we're continuing monitoring and slicing and dicing it in different ways. And as we have information or insights to you, we'll share it. And without trying to throw a whole bunch of cold water on our mind, it was slightly better or a step-function change better, but good to see.

Terrific. And maybe for a quick follow-up for Dev. I think it builds off maybe Tyler's question at the top of the Q&A, but you had cited that some customers are thinking about their workloads more holistically and even looking to run AI workloads on-prem. How much of that do you think is just a function of customers still trying to figure out how to optimize for latency and cost or is this more a demonstration of we really are in the early phases of the exploratory phase versus going into production? Is there any way to coarse that out or is the two not necessarily connected? Thank you.

No, I think it's a bit of both. Some customers are very committed to running a significant portion of their operations on-premises. Therefore, if they decide to build an AI workload, it has to be executed on-prem, which requires access to GPUs, and they are obtaining those. Other customers are renting GPUs from cloud providers to develop their own AI workloads. I believe we are still in the early stages of this journey. Customers are still learning and experimenting. More applications are moving into production, and as I mentioned earlier, we have thousands of AI workloads running on MongoDB, but only a small percentage have shown significant product market fit. So, initial traction is somewhat limited. However, as people become more adept with AI and as the technology evolves and gains utility, we will see applications begin to flourish. I find it amusing that today, senior leaders seem to be more focused on the chips they are using rather than the applications they are developing, which indicates we are still in the very early phases of this major platform shift.

Great point. Thank you again, guys.

Thanks, Mike.

Thank you. And one moment for the next question. Our next question will be coming from the line of Eric Heath of KeyBanc. Your line is open.

Hi, thanks for taking the question. Dev, Michael, it sounds like the takeaway from the call is a greater focus on EA and on enterprise. So should we structurally rethink the EA business differently and think of this more as a healthy double-digit growth business going forward for the foreseeable future? And then if I could just ask a follow-up question separate to that. But Michael, I understand that it's still early to identify the fiscal '25 cohort of workloads, but just curious at a high-level if they look and feel like of higher quality than the fiscal '24 cohort of workloads.

Yeah, I mean, I would say, I mean, we are very committed to our run anywhere strategy. And as I said, we are first investing in community where for many customers is the first way they experience MongoDB. And we want them to have the full experience of integrating search and vector search into our core product. And so they can out of the gate really start building applications. That will then transition to building those capabilities into EA. So, we are clearly investing in the EA product. But Atlas is still a big, big part of our business and a big, big part of our growth engine, and we typically launch new features on Atlas. Because of the capabilities we already have, the fact it's multi-cloud makes it a very, very compelling offering for many customers.

Yeah. And I think in terms of the workloads, I do think it's early. Just as a reminder for folks, they tend to start small, although grow quickly. I think the only other thing that I can add is we've been pretty consistent in that we've been pleased with the new business that we've done. But we need some time to let the cohorts play out as we track them. But I think, like I said, we've been happy with the new business that we were winning.

Thank you. And one moment for the next question. And our next question will be coming from the line of William Power of Baird. Your line is open.

You there, William?

Sorry, yeah. Thank you. Dev, you had some encouraging comments on relational migrator. I wonder if you could just touch on what you think is driving the higher interest here. I mean, it sounds like AI is contributing and helping, but it'd be great to get some more color there because that still feels like obviously a meaningful long-term opportunity. And then maybe the second part of the question for Dev or Michael, just be great to get any other framework around the professional services investments. Any way to kind of think about quantification and timing of that?

We are excited about the opportunity to target legacy applications because several factors are converging. The costs associated with supporting and managing these legacy apps are increasing significantly. In regulated industries, clients are facing pressure from regulators regarding the systemic risks of maintaining these outdated systems, which means they can no longer postpone upgrades. Additionally, some vendors are phasing out their support, forcing companies to consider migrating to more modern platforms. As businesses recognize the importance of leveraging proprietary data for competitive advantage, there is a growing desire to modernize and access their data more easily. Furthermore, many of the individuals who created these applications are retiring or leaving, which adds to the risks for these companies. Given these circumstances, customers are keen to find a safe and efficient way to transition away from these legacy applications. We can assist them in moving data and converting relational schemas to document schemas. Previously, the most challenging aspect was rewriting the applications, but with the emergence of GenAI, this process can be significantly accelerated. GenAI can analyze existing code, reverse engineer tests to understand the code’s functionality, and help generate new code, which can then be tested to ensure it produces the same outcomes as the old version. This reduction in time and effort has sparked substantial interest from clients, especially since simply switching from one relational application to another doesn't represent real modernization. Transitioning to MongoDB offers a more advanced platform that is agile, flexible, high-performing, and scalable for future needs. We are enthusiastic about this potential. Although we are in the early stages, several pilot projects have gone well, and we are currently collaborating with customers on migration plans. However, this is a complex process that will take time, as clients are not just looking to replace less critical applications, but rather some of their most essential systems. This focus on core applications is particularly encouraging. While this transition will require time and investment, we are committed and believe it will drive meaningful long-term growth.

Yeah, Will. And to the last part of your question on the professional services investment, we're really building out that capacity in order to meet the demand that we're seeing relative to the opportunity. We're calling it in particular because it has a gross margin impact because that's where that will typically show up. And then maybe the last thing, and it's probably obvious, but just to sort of underscore it is the reason we're doing this though is for the ARR, right, to drive the new workloads, the additional workloads over to MongoDB as part of that migration. Over time, as we've talked about before, we hope and expect to be able to leverage technology more and more, but at least initially and into the medium-term, there's going to be a healthy human/services component to that. Just wanted to sort of effectively telegraph that out to folks.

That's helpful. Thank you.

Thanks, Will.

Thank you. One moment for the next question. And our next question will be coming from the line of Rudy Kessinger of DA Davidson. Please go ahead.

Hey guys, thanks for squeezing me in here. I believe last quarter you said consumption growth slightly ahead of expectations. And while down slower year-over-year growth versus Q2 last year, the year-over-year growth did improve from Q1. I guess I'm curious for Q3, could you make a comment in that same regard? Obviously, slower on a year-over-year basis than Q3 last year, but was it stable with year-over-year consumption growth in Q2 or better or worse?

Yeah. Rudy, thanks for the question. We haven't specifically called that out relative to Q2. We did see a lower year-over-year growth, as we called out. We did see a seasonal rebound. Usually, Q3 is stronger than Q2, and we talked about how that was smaller than in the prior year. So hopefully, that will help you all triangulate.

Okay. And then just a quick follow-up. I believe it was on your Q4 call back in March. At that point, Dev, you said it would be at least another year until AI applications are being deployed at scale. It sounds like the commentary that some early large workloads, but out of the thousands, just not many that are at large scale. I guess, is your expectation now that maybe it's still at least another year until we're seeing broad AI application rollouts at scale?

I think much of this relates to the current state of AI research and development. For instance, we currently lack a robust model tailored for smartphones, and existing computers do not possess enough power to operate complex models effectively. As a result, there aren't many successful consumer applications aside from a few like ChatGPT or Claude. We're not witnessing the surge of hundreds of apps that characterized the early days of the internet, the cloud, or mobile technology. We are still in the initial stages of AI development. While numerous AI applications are being created, many have basic functionalities. However, I believe this will evolve over time, and I am confident it will change. I just can't specify when that will occur. In areas where we do see apps gaining traction, they are growing rapidly, and we have many on our platform, though very few have significant impact.

Thank you. And that concludes today's Q&A session. I would like to go ahead and turn the call back over to Dev for closing remarks. Please go ahead.

Thank you, everyone. We're really pleased with our Q3 results, featuring strong new business performance and revenue that exceeded expectations across both Atlas and EA. We are investing to expand our enterprise channel where we see the biggest opportunity to make MongoDB a standard and achieve the best returns on our go-to-market investments. Looking ahead, we feel encouraged by our progress in accelerating legacy application modernization with AI and establishing ourselves as a standard in the emerging AI tech stack for new AI applications. Lastly, I want to thank Michael for his contributions over the past 10 years and wish him well. Thank you, everyone, and we'll talk to you soon.

Thank you for participating in today's conference call. You may all disconnect now.