ReadySet, a company providing database infrastructure to help developers build real-time applications, today announced that it raised $24 million in a series A funding round led by Index Ventures with participation from Amplify Partners. Several angel investors also contributed, bringing ReadySet’s total raised to $28.9 million — building on a previously-undisclosed $4.9 million seed round.
According to cofounder and CEO Alana Marzoev, ReadySet is tackling a major challenge in the enterprise having to do with delivering dynamic content while servicing large, distributed customers. The current standard practice is to build custom query caching systems, but Marzoev claims that this can slow down engineering teams, drive up costs and cause outages at inopportune times.
“Rather than rebuilding these same broken systems, developers need solutions that slot into their existing infrastructure and achieve limitless read scaling,” Marzoev told TC via email earlier this month. “With ReadySet, we aim to make the process of globally caching … query results as streamlined and automated as caching images in a content delivery system.”
ReadySet’s product has its origins in research that Marzoev and the company’s second cofounder, Jon Gjengset, did at MIT while pursuing their doctorates. Marzoev was previously a cloud infrastructure researcher at Microsoft, where she worked on cloud networking and storage infrastructure technologies, while Gjengset was a senior software development engineer at Amazon Web Services.
Together at MIT, Marzoev and Gjengset spearheaded an open source project called Noria, a streaming data-flow system designed to act as a fast storage backend for web apps. After the project gained traction on GitHub, the two decided to refine it and bring it to market as a managed service: ReadySet.
“The traditional databases that back the world’s most popular applications catch fire when dealing with large datasets, complicated queries, or high request volumes — in other words, at the worst possible time, when the product starts gaining traction,” Marzoev said. “To handle the infrastructure challenges that come along with this growth, companies scramble to hire teams of engineers with specialized skill sets who can help build and maintain custom, in-house solutions. ReadySet circumvents this problem by providing limitless read scaling without requiring any code changes to integrate it into the application.”
Robustifying the query backend
To take a step back, enterprises leverage several different kinds of databases to store, serve and analyze their app data. By far the most common type are relational databases, which provide access to data points that are related to one another — as the name implies. Programming languages called relational query languages use algebra to interpret requests about data and then instruct a database management system, or DBMS, to execute the requests.
A relational query or request, then, is a request asked about data contained in two or more tables in a relational database.
ReadySet acts like a database, but precomputes and caches relational query results so that reads of data in the actual databases remain fast. The platform keeps cached results up-to-date as the underlying data, stored in persistent “base tables,” change.
“At its core, ReadySet accelerates queries via a novel … caching engine that keeps cached state up-to-date automatically while supporting millions of reads per second with sub-millisecond latencies on a single node,” the company claims in its press materials. “ReadySet slots in front of existing relational databases and … can be integrated into existing [apps] without code changes.”
Alternatives to ReadySet’s platform exist in the form of Materialize, an open source project akin to Noria, and key-value stores like Memcached, Redis and Amazon ElastiCache. (Key-value stores record data in a “key-value” format where data is fetched by a unique key or keys, optimized for reading and writing that data.) But Lenny Pruss, a general partner at Amplify Partners, argues that both competing systems and key-value stores aren’t “feature complete” and require writing custom logic to get them to work for relational database applications, in contrast to ReadySet.
“Over the last decade, we’ve seen applications become more dynamic, real-time and global, all the while innovation in data access technologies has lagged behind,” Pruss told TC via email. “This has resulted in a heavy burden placed on engineering teams struggling to make do with overly complex caching and/or database sharding architectures. We believe ReadySet offers a disruptive new approach to not only speed up application performance but free engineering teams from toil.”
Future expansion plans
ReadySet is currently pre-revenue stage, but the company is collaborating with a small number of potential customers as design partners. As ReadySet works toward a more generally available product, funded in part by the series A proceeds, brands will be able to sign up for early access, Marzoev says.
If surveys are any indication, there’s enterprise-sized appetite for solutions like ReadySet. In a 2021 survey by 451 Research, commissioned by Immuta, 55% of companies reported that their data is often stale or out-of-date by the time it’s consumed or analyzed. A separate poll by Dimensional Research for Fivetran suggests that the majority of companies, meanwhile, experience problems with pipelines to access their data breaking more than once per month.
“Our short-term aim is to drastically improve the speed and usability of caching. Our long-term vision is that caching is something that no developer ever has to think about again. You just hook ReadySet up to your application and it handles understanding what needs to be cached and when,” Marzoev added. “Internet user growth set records in the pandemic, but database performance has stayed the same. We’re responding to tremendous demand among enterprise and fast growing companies that are looking for a way to meet rapid growth goals with scalable caching technology.”