Now gives bi-directional integration with Iceberg, quicker vector search, enhanced full-text search, autoscaling and a “deliver your personal cloud” deployment
SingleStore, the real-time information platform so that you can transact, analyze and contextualize, introduced a bi-directional integration with Apache Iceberg that opens up a world of alternatives for constructing clever functions on information lakehouses. This integration addresses the important problem confronted by enterprises the place an estimated 90% of information stays “frozen” in lakehouses and is unusable for powering interactive functions, analytics or AI.
Apache Iceberg is the de facto customary for information lakes for storing large datasets, providing price effectivity for querying large-scale datasets. Nevertheless, enterprises wrestle to leverage this information on account of complicated, pricey processes required to ‘thaw’ it, together with in depth ETL workflows and compute-intensive Spark jobs.
SingleStore’s new integration is meant to unravel exactly that, offering low-latency ingestion, bi-directional information move and real-time software efficiency at decrease prices for contemporary clever functions and analytics. This providing is obtainable in public preview right this moment and clients will quickly be capable of create exterior tables in SingleStore based mostly on Iceberg information, construct projections on these tables and make the most of the velocity and efficiency of SingleStore on frozen information.
SingleStore’s clients, a lot of whom use Apache Iceberg for his or her information lakes and business lakehouses, have expressed a powerful curiosity in an answer that permits quick interactive functions and low-latency analytics utilizing their huge, untapped Iceberg information.
“Our imaginative and prescient has all the time been to offer one single information retailer for all corporations to have the ability to make the most of velocity, scale and ease,” mentioned Raj Verma, CEO, SingleStore. “Our information platform is designed to unlock all sorts of enterprise information — together with information that’s frozen in information lakes — to allow our clients to construct trendy clever apps. With this launch, we imagine we’re enabling a good portion of the market that right this moment can not construct real-time trendy functions on information saved in information lakes.”
Aside from the Iceberg integration, SingleStore additionally introduced a slew of different new options and product enhancements that allow constructing enterprise-grade clever functions, together with:
● Sooner vector search. Vector search with HNSW is 40% quicker in comparison with the sooner launch. SingleStore’s IVF Flat index is demonstrated to be between 47 to 100x quicker than pgvector, and vector index construct occasions are actually 2-3x quicker than Milvus and pgvector. Extra vector search options embrace vary searches and filters (in public preview) to assist enterprises simply construct and scale generative AI functions.
● Enhanced full-text search. SingleStore has supported full-text seek for a variety of years. With its June 2024 launch, the corporate is saying the final availability of strong new capabilities in full-text search together with improved relevance scoring, phonetic similarity, fuzzy matching and keyword-proximity-based rating. All this implies organizations can now simplify their information architectures by eliminating the necessity for extra specialty databases to construct generative AI and real-time functions.
● Autoscaling. With compute workspaces and a distributed structure, SingleStore is a extremely scalable information platform. With Autoscaling in its newest launch (in public preview) clients guarantee software efficiency by scaling compute assets up or down routinely in seconds to regulate to unpredictable workloads — whereas avoiding billing surprises.
● Helios® — Deliver Your Personal Cloud (BYOC) deployment. For a lot of clients, their information has to stay in their very own digital non-public cloud (VPC), beneath their very own safety and governance insurance policies due to regulatory necessities or contractual obligations. In these circumstances clients needed to deploy SingleStore in a self-managed mannequin, managing operations themselves. With this announcement, SingleStore is introducing a fully-managed cloud providing (Helios — BYOC), that lets clients benefit from the ease of administration and elastic scalability of the platform, however inside their very own digital non-public cloud. This providing is now obtainable in non-public preview in AWS, and clients can run SingleStore in their very own tenants whereas complying with information residency and governance insurance policies.
“SingleStore’s information platform is indispensable for analyzing our information and delivering real-time insights to our clients,” mentioned Mauricio Aristizabal, Principal Knowledge Architect, impression.com, and Founder, Giist Social. “This newest replace continues to push the boundaries of what we will accomplish by harnessing all of our enterprise information in actual time. We’re wanting ahead to leveraging the newest cutting-edge options to proceed rising our and our purchasers’ enterprise.”
“Our new options and capabilities advance SingleStore’s mission of providing a real-time information platform for the subsequent wave of generative AI and information functions,” mentioned Nadeem Asghar, Head of Engineering at SingleStore. “The mixing with Iceberg, plus enhancing our platform’s scaling and processing capabilities will permit customers to entry and harness their information in actual time, and construct clever apps.”
SingleStore stays on the forefront of generative AI innovation. The corporate’s product innovation has led to high-momentum development in clients, pushing it to surpass $100M in ARR on the finish of 2023. SingleStore was acknowledged with 5 2024 TrustRadius High Rated Awards, essentially the most the corporate has ever gained in a single 12 months. 4 awards had been gained for the third consecutive 12 months, whereas the popularity within the Vector Database class was a primary.
Join the free insideAI Information e-newsletter.
Be a part of us on Twitter: https://twitter.com/InsideBigData1
Be a part of us on LinkedIn: https://www.linkedin.com/firm/insidebigdata/
Be a part of us on Fb: https://www.fb.com/insideBIGDATANOW