A lot has happened since Dev Ittycheria took the helm of MongoDB, the $26 billion database company he led as president and CEO since September 2014. Ittycheria is where he moved MongoDB to the cloud, ran it through its IPO, oversaw its transition from open source, and launched it. He became a venture capital arm and grew his client base from a few hundred to nearly 50,000.
“When I joined, it wasn't clear that people would trust us to be a truly mission-critical technology,” Itticheria told TechCrunch. “When I joined, we had about $30 million in revenue. Now we're close to $2 billion in revenue.”
However, it wasn't all peaches and cream. Five months ago, MongoDB suffered a security breach that, while relatively contained, temporarily put its reputation at risk in an industry where reputation matters most.
Add to that the whirlwind AI revolution that has engulfed almost every industry, and there was a lot to discuss when TechCrunch sat down with Ittycheria at MongoDB's new London office, which opened in Blackfriars last year.
MongoDB's London office. Image credit: Paul Sawers / TechCrunch
vector hug
Databases have come a long way since IBM and Oracle first popularized relational databases more than half a century ago. The rise of the Internet created a demand for flexible, scalable, and cost-effective data storage and processing, paving the way for businesses like MongoDB to flourish.
Founded in 2007 by three veterans from online ad tech company DoubleClick (which Google acquired for $3.1 billion), MongoDB was originally called 10Gen before being rebranded to its flagship product name six years later. I did. Since then, his NoSQL database has emerged as one of the best, helping businesses store and manage large amounts of data.
Prior to joining MongoDB, Ittycheria founded and left a server automation company called BladeLogic in 2008 for $900 million, and has since held various board member and investor roles, including 16 months at Greylock. He joined MongoDB as President and CEO. It's been 10 years since now. Itticheria replaced Max Silleson, who stepped down after just 18 months in office due to family reasons.
Built on a document-oriented model, MongoDB has thrived on the explosion of mobile and web applications that rely on flexible, dynamic data structures. The current wave of artificial intelligence is driving similar changes, with vector databases becoming the hot new thing in town.
Like NoSQL, vector databases are also specialized for unstructured data types (images, videos, social media posts, etc.), but are especially suited for large-scale language models (LLM) and generative AI. This is due to the way we store and process data in the form of vector embeddings, which store data spatially by association, converting it into a numerical representation that captures the relationships between different data points. This makes it easier to capture semantically similar data, allowing AI to better understand the context and meaning within a conversation.
Over the past few years, we've seen a number of startups dedicated to vector databases, but established companies like Elastic, Redis, OpenSearch, Cassandra, and Oracle are also starting to adopt vectors. Cloud hyperscalers such as Microsoft, Amazon, and Google are also increasing their support for vector searches.
Last June, MongoDB also introduced vector search to its flagship database-as-a-service product, Atlas. This shows the company is preparing for the upcoming AI tsunami. This mimics other historical trends in which single-function databases (e.g., time series) emerged that had utility as standalone solutions, but were better integrated into larger multipurpose database stacks. There is a possibility. This is exactly why MongoDB introduced support for time series databases several years ago, and why we will do the same with vector.
“A lot of these companies hide functionality in their products,” Itticheria said of the new wave of dedicated Vector products. “We build it into the platform and that's the value. Rather, we use a standalone vector database and then use OLTP. [online transaction processing] Databases and search databases, all three in one platform makes life much easier for developers and architects. ”
The idea is that a database provider that takes a multifaceted approach will combine all the data into one place, making it easier for developers to work with it.
“There are probably 17 different databases and probably 300 vendors,” Ittycheria says. “There is no customer on this planet that needs 17 different databases. The complexity of creating these different technologies and the costs to learn, support, and manage them are enormous. This tax also hinders innovation.
MongoDB developer Ittycheria. Image credit: MongoDB
Too much hype
Despite the readiness, Itticheria believes there is too much hype around AI, at least for now.
“AI hasn't changed my life,” he says. “Yes, the assistant may help me write better emails, but it hasn't fundamentally changed my life. The Internet, on the other hand, has completely changed my life.”
In theory, despite all the fuss, it will take time for AI to permeate our daily lives, and when it does, the applications that integrate it and the businesses that build on it It is said that this will be done through.
“I think when you introduce new technology, you create value first at the bottom,” Itticheria said. “Obviously, Nvidia is making good money, and OpenAI is the most talked about company since they launched ChatGPT. But the real value is when people build applications on top of those technologies. That's our business: helping people build applications.”
For now, it's all about “simple apps,” as Ittycheria puts it. This includes chatbots for customer service. This is something MongoDB itself does internally with his CoachGTM, which leverages MongoDB's vector search to provide sales and customer teams with instant information about the product. In a sense, we are now at the “calculator app” stage that the iPhone reached about 20 years ago when the concept of the App Store hit the masses.
“The truly sophisticated ones are [AI] The app will be able to use real-time data and make real-time decisions about real-time events,” Itticheria said. “Maybe something is going on in the stock market, or it might be time to buy or sell, or it might be time to hedge. From there, more sophisticated apps emerge that can embed real-time data with all the inference. I think I'll start.”
SaaS Pass
One of the biggest developments during Ittycheria's tenure has been the move away from a self-deployment model where customers host MongoDB themselves and the company sells features and services. With the launch of his Atlas in 2016, MongoDB embarked on the familiar path of his SaaS, where companies charge a fee to remove all the complexity of self-hosting. At the time of his IPO the following year, Atlas accounted for his 2% of MongoDB's revenue, but now that number is close to his 70%.
“It's growing very quickly and we've really built that business as a publicly traded company,” Itticheria said. “What the popularity of Atlas has shown is that people are comfortable using infrastructure as a service. will be able to delegate. So you can focus on building applications that truly transform your business. ”
Another major development was the transition from the open source AGPL license to the source-available SSPL (Server Side Public License) a year after MongoDB went public. In some ways, this was a harbinger of things to come, with countless infrastructure companies continuing to abandon their open source credentials to prevent cloud giants (like Amazon) from selling proprietary versions of their services for nothing in return. .
“We are very happy that [the license change]” Itticheria said. “The reality is that even though it's open source, 99.9% of the development is done by our own staff and it's not like the community contributes code. This is not a simple, trivial application; It's a complex piece of code, and we have to spend a lot of money and hire talented senior people. We spend all this money to build this product, and then someone gives us that free product. We thought it wasn't fair to take it and monetize it and not give us anything back, which was quite controversial in 2018, but looking back, our business has grown even faster. Masu.”
And it grew. Like nearly all technology companies, MongoDB's valuation soared during the pandemic, reaching an all-time high of $39 billion at the end of 2021, but then plummeting to less than $10 billion within a year. Did. This is about the same as pre-pandemic numbers.
However, MongoDB's stock price has risen over the past 18 months since then, reaching $35 billion just a few months ago, but has now fallen again to around $26 billion. This is the volatile nature of the stock market. However, given the company's relatively modest valuation of $1.8 billion at the end of its first day of trading in 2017, MongoDB has performed fairly well for public market investors.
Dev Ittycheria and MongoDB colleagues at their 2017 IPO. Image credit: MongoDB
However, four months ago, MongoDB disclosed a data breach that exposed “some customer account metadata and contact information.” This included phishing attacks via third-party enterprise tools (which Ittycheria has not confirmed). This caused MongoDB's stock price to fall by 3%, but in the months that followed, MongoDB's valuation returned to a two-year high. This highlighted how small an impact this breach had on internal operations compared to high-profile data breaches such as Equifax and Target, which devastated businesses and forced senior executives to resign.
While the scope of the MongoDB cybersecurity incident was significantly smaller, what stood out was how quickly it all went away. The incident was reported in several news outlets (including TechCrunch), but the story disappeared into the foggy ruins of time as quickly as it arrived.
“Part of the reason is we've been very transparent,” Itticheria said. “The last thing you want to do is hide information and appear to be misrepresenting it. Many banks store highly sensitive information on data platforms. We have a lot of companies that we own. So what's important for us is to make sure that the architecture is robust and healthy. And this has forced us to double down on it again. We never say that, but we do everything we can to avoid being hacked.”
nothing adventurous
It's not uncommon for the biggest technology companies to launch their own investment vehicles, as we've seen over the years with Alphabet (which has several investment subsidiaries), Microsoft, Amazon, and Salesforce all Embracing the startup fraternity. But a new wave of enterprise ventures are also entering the fray, including Slack, Workday, Twilio, Zoom, HubSpot, and Okta.
In 2022, it was MongoDB's turn to launch such a fund, and in the two years since then, MongoDB Ventures has invested in about eight companies.
“This is because we build deeper relationships. We work in an ecosystem made up of large and small companies,” Itticheria said. “When we find a small company that we think would be interesting to work with, we say, 'We'd like the chance to invest in you,' so that creates extra value. We get some of that value.” You are also the beneficiary of this.”
Ittycheria emphasizes that MongoDB is putting investments on the back burner because its corporate development team has only a few people who are primarily focused on venture funds. We also typically invest in parallel with other VCs, as we did with our first investment in 2021 (before the fund's official launch), including Apollo GraphQL's $130 million investment. It quietly participated in the Series D round with Insight Partners, Andreessen Horowitz, and others.
“We will always be in the minority, we will not sit on the board and we will not set the conditions,” Itticheria said. “But the reason startups are interested in us is because they want to leverage the MongoDB brand. We have thousands of people in the field. [startups] Take advantage of our distribution channels. ”