As we start the new year, we wanted to share a few highlights of customer, product and developer community news from SingleStore in 2021. It has useful information for learning about distributed SQL databases, data-intensive applications and the outcomes developers are achieving, so you may want to bookmark this post. To learn more about SingleStore company milestones this past year, check out our CEO Raj Verma’s post.
Customers Delivering Data-Intensive Applications
In 2021, we added the most new customers ever in the company’s history, expanded use significantly with existing customers and celebrated some customers’ IPOs. The new customers are innovators in industries like fintech, gaming, media, communications, marketing technology and cybersecurity, and are providing modern data experiences to their customers. The common challenge they faced was how to cost-effectively solve for data intensity. Their data-intensive use cases — like operational analytics, in-product analytics, portfolio analytics, fraud prevention, gaming telemetry, real-time esports analytics, real-time marketing analytics and algo trading — all demanded a new approach to the cloud data stack that solved for the crucial time-sensitive moments and digital experiences their own SaaS products demanded.
The common obstacles they needed to solve were data bottlenecks across several dimensions: data size, speed of ingestion, latency, query complexity and concurrency. When an application exhibits two or more of these challenges simultaneously, we call these data-intensive applications. SingleStore is the industry’s first cloud database built for data-intensive applications.
We also saw an increase in adoption of hybrid cloud deployments for SingleStore. Our fully-managed cloud database service (DBaaS) is now in its third year of availability on AWS, Azure and GCP — but in 2021, we introduced a hybrid cloud deployment option. SingleStore can now be deployed in a hybrid cloud configuration where you get the full capabilities of the product (e.g., separation of storage and compute with Unlimited Storage) but separate the administration from the data. This allows you to have the best of both on-premises and multi-cloud worlds. The hybrid cloud deployment allows you to own the data plane in any Kubernetes environment including your own data center, AWS VPC, Azure VPC or GCP VPC, and connect to the SingleStore control plane. The data plane contains your data and the nodes of the database, while the administrative control plane is hosted by SingleStore. The benefit of this is that customers can keep their data in their existing environments. Customers in diverse verticals such as energy, sharing economy services, accounting and financial services are now running their SingleStore databases in the hybrid cloud, or are in the early stages of adoption.
We are delighted that a few customers agreed to share their data-intensive success stories publicly with the SingleStore community this year, especially since so many view SingleStore as the secret weapon in their strategic cloud data stack. We love hearing how our customers are making use of the unique properties of SingleStore’s modern cloud database in their organizations, and we hope you find these kinds of stories informative and inspirational:
- IEX Cloud - Serving more than 1.2 billion API calls daily to over 200 financial datasets, IEX Cloud is the financial data platform built to connect developers and data creators. “SingleStore enables us to do monitoring and analysis in the same system that houses our historical data, and this creates enormous efficiencies for us,” said Josh Blackburn, Chief Technology Officer of IEX Cloud. Their data prep and ETL process execution time dropped from days to just minutes for data updates, and the API’s average response time is now only 8ms for a working set of 100s of TBs across real-time and historical data. After considering more than 10 databases, including Clickhouse, CockroachDB, Yugabyte and Google BigQuery, the team selected SingleStore because, as Josh stated in this webinar, “It met all the requirements we had for writes, massive analytical data, and web-scale traffic.”
- Factors.ai - The death of the dashboard has been widely reported and Factors.ai may be one of the drivers of that trend. Factors.ai offers an Amazon Alexa-like interface to B2B marketers so they can ask a question and receive a set of specific, focused recommendations based on their own data, rather than having to inspect hundreds of dashboards. They were hitting a wall with their cloud-hosted PostgreSQL database. CTO and co-founder Aravind Murthyshared with SingleStore, “We went from taking 20-30 minutes to run a query on 50 million records to 10 seconds. That's a 180x faster query response.” And Chief Product Officer Praveen Dasr shared, “...our clients can plan out a campaign, generate an analytics report, create a user segment, send it to their targeting system, and take it live in one hour. That’s about 32x faster than before and it’s the sort of competitively differentiating agility we bring to marketing,”
- Fathom Analytics - One of the most-widely shared stories of the year was Fathom, due in no small part to its loquacious and gracious co-founder and CTO, Jack Ellis. The company provides website analytics which is fully privacy law-compliant. When faced with scaling their cloud-hosted MySQL database to serve quickly-growing analytical queries, Jack discovered that not only could SingleStore solve his analytic query performance and user concurrency bottlenecks, but it also could serve as the application database outright replacing MySQL for OLTP, it could replace Redis as the caching tier and replace AWS DynamoDB as the key-value reference store. This greatly simplified the data stack and reduced the DevOps burden significantly resulting in more than a 60% reduction in database total cost of ownership and a 1000x increase in query performance. Building in the open, he shared his unfolding journey with the community through his blogs, t“Building the World’s Fastest Website Analytics,” “Making the World’s Fastest Analytics Even Faster” and “Why We Ditched DynamoDB.”
We also celebrated the IPOs of several SingleStore customers in 2021. We are grateful to be a part of their success and their modern cloud data stacks:
Launching Mission-Critical Capabilities
The product innovations developed in 2021 have been remarkable. The development focus for the year was to advance the product to better serve enterprises’ mission-critical scenarios for their applications. SingleStore has been well-known as a fast database for real-time analytics since its early days as MemSQL. But in the last few years, the product has improved its ability to store and process diverse streaming data for low-latency analytical workloads, while simultaneously expanding to capabilities heretofore only seen in cloud data warehouses and operational databases. This expansion to better support a widening array of workloads is part of a larger data management market trend that SingleStore is leading — and disrupting the old guard of the last 50 years of on-premises databases, as well as the last 15 years of cloud databases.
Multiple industry analysts have identified this technology unification trend from different vantage points. Stephen O’Grady from RedMonk remarked that the market is searching for capabilities now in a new kind of modern general purpose database. Rick van der Lans has explored the question of whether unified databases make polyglot persistence irrelevant. Mike Gualtieri and Noel Yuhanna from Forrester have been covering two specific parts of the unification trend with the Translytical Data Platform, combining operational and analytical workloads, and the Multimodel Data Platform, which combines diverse data models and data types in a data platform. And Gartner continued its coverage of Cloud Database Management Systems in the Gartner Magic Quadrant released in December, covering both operational and analytics capabilities across a single cloud database market.
They used to call this transactional and analytical unification HTAP, but have since redefined it to cover an augmented set of capabilities that includes machine learning. SingleStore was recognized in this year’s Gartner MQ for cloud database systems for the first time. Some folks in the industry talk about this database unification trend in the future tense but at SingleStore, the future is already here, to riff on William Gibson.
SingleStore has developed unique innovations to achieve simplicity, power, and cloud resource and cost efficiency in a single general-purpose modern cloud database while avoiding the need for difficult tradeoffs for diverse application workloads. The company was awarded patent # 11,068,454 in July for one of those innovations, Universal Storage, which was declared feature-complete last summer.
Here are some more of the mission-critical enterprise SingleStore product capabilities launched in 2021:
- System of Record capabilities
- Lock-free Backup for online backup without pauses
- Multi-AZ Failover
- Resource Governance
- Achieved 4 nines (99.99%) SLA
- Ease of Use
- Faster Cluster Creation
That’s a big list, so I encourage you to check out the summer announcement and its release notes, as well as the fall announcement and its release notes. Of course, if you like to sit back, watch and listen,check out the accompanying webinars.
Building the Developer Community
TikTok much? If not, you may find that’s a channel that is expanding quite a lot. You’ll find our Developer Advocate Joe Karlsson here, there and everywhere. It’s just one additional channel we expanded into to meet, educate, support and learn from our growing developer community where they already are. We expanded the Community and Developer Relations teams, and began collaborating with Engineering’s Launch Pad team. Here are just a few of the milestones these teams have achieved together this past year:
- Contributed many helpful how-to articles for getting started, using SingleStore in various multimodal use cases, the recipe for building a modern database, using Streamlit with SingleStore, using SingleStore as a feature store, and much more
- Livestreamed Tech Talks for getting started, and for specific use cases like streaming data in parcel logistics
- Contributed to advancing the state of the art for WASM and WASI through the Cloud Native Computing Foundation, CNCF
- Established a new GitHub organization called SingleStore Labs focused on the useful code and frameworks like geospatial track and trace of airplanes, and serverless CloudFlare Workers with SingleStore
We are giddy with excitement about what’s in store for 2022. Across community, customer and product we have big plans this year and we will be sharing a lot over coming months. We are on a mission to bring together modern data to make it work for you in the moments that matter to your customers. We simplify data architectures so you can avoid the burden of stitching multiple datastores together, and maintain them to meet your data-intensive challenges. We were built for this moment: SingleStore is the industry’s first cloud database built for data-intensive applications, and it’s a great time to join us. You can do that by joining our developer community or the company…or both! We are rapidly growing in Engineering, Marketing, Sales and other areas. You’ll find all of the open positions on our Careers page. Be part of the data revolution with SingleStore!