April 2025 Summaries

The SingleStore team has rolled out cron-based scheduling for Notebooks in their platform, allowing users to schedule workloads with pinpoint accuracy using industry-standard cron syntax. This feature eliminates execution drift, offers precision timing, and supports complex scheduling patterns. The new feature is designed to be intuitive for both cron experts and newcomers, with a user interface that validates cron expressions in real time. It's particularly valuable for data analysts, engineers, and scientists who need precise timing for their workloads. The feature seamlessly integrates with existing Notebook execution features and provides comprehensive documentation to help users get started. The SingleStore team is committed to ongoing enhancements and welcomes feedback from users as they start using the new feature.

With SingleStore Helios, users can expect a faster, more efficient experience through significant performance optimizations to the Command + K feature. This update refactors rendering logic and removes redundant React hooks, resulting in a 96% reduction in SQL queries and a 97% reduction in sequential execution time. Parallel execution times also see an improvement of over 84%. The enhancements provide users with at least 5x faster search experience, along with streamlined UI, optimized keyboard navigation, advanced fuzzy matching, sectional search capability, and instantaneous search results. Ongoing optimization and enhancement are planned, with more updates already underway.

SingleStore Flow is a tool designed to simplify the process of migrating data from various sources such as Oracle, SQL Server, MySQL, PostgreSQL and Snowflake into SingleStore databases in real-time. It offers a seamless experience through its Ingest and XL Ingest components, which handle smaller and larger table transfers respectively, ensuring efficient and speedy data transfer. Flow provides lightning-fast migration, no-code data replication, multi-threaded parallel loading and time-series data capabilities, making it an attractive option for enterprise data integration. Its intuitive GUI and automatic network catch-up features further enhance its usability. By bridging data silos and enabling robust analysis, machine learning and operational reporting, SingleStore Flow offers a solution to unite disparate data sets, reducing the need for cumbersome scripts and manual processes. The tool's easy installation and economical pricing make it an attractive choice for those looking to streamline their data integration efforts.

To use the SingleStore MCP server with Claude, users must first install the Claude Desktop application on their desktop and set up a local SingleStore MCP server by cloning its GitHub repository and running the server source code. Once the server is set up, users can connect it to the Claude Desktop application by making settings changes in the application's configuration file. The connection requires an API key generated from the SingleStore Portal, which can be obtained by navigating to the "API Keys" menu item under the "Configuration" section. Users can also access database credentials if desired. With the server connected and configured, users can start using the SingleStore MCP server in a Claude chat to discover available tools and perform various tasks such as analyzing data.

We're excited to announce the launch of our new Notebook Version History feature, a powerful tool designed for data science projects that gives you peace of mind while working on your notebooks. This git-like experience integrates directly into our notebook interface, allowing you to track changes, compare versions and easily revert to previous states of your work. The feature eliminates hassle by automatically tracking changes and giving you the ability to view a chronological list of versions, side-by-side comparison, restore or duplicate with one click, bookmark milestones, streamline version tracking, and more. To get started, simply open any notebook, look for the "Version History" clock icon in the toolbar, and click it to view your saved versions. We're already exploring several exciting enhancements based on user feedback, including expanding to SQL files and beyond, database time travel integration, semantic diffing and feature-rich diff viewer, and enhanced RBAC security controls.

SingleStore has developed a database that combines the benefits of SQL and NoSQL, allowing users to store structured data in JSON format and still achieve fast analytics and query performance. This is achieved through the use of columnarization, vectorized query engines, and native support for MongoDB-compatible APIs, making it possible to run analytical queries on JSON data without sacrificing ease of use or requiring schema changes. With SingleStore, users can now handle both structured and semi-structured data in a single database, supporting AI applications that rely on both SQL and NoSQL data. This allows for consistent, predictable performance and eliminates the need to choose between SQL and NoSQL.

SingleStore Helios BYOC regions offer customers a flexible and secure deployment option on AWS, allowing them to maintain control over their cloud infrastructure while benefiting from SingleStore's expertise in managing the underlying database service. This hybrid approach combines the benefits of self-managed deployments with the operational efficiency of cloud services, providing customers with the flexibility to deploy their SingleStore instances within their own AWS account and VPC. BYOC regions prioritize customer sovereignty, security, and isolation, ensuring that customer data remains private and hosted exclusively within the customer's AWS account. The managed Kubernetes experience ensures seamless scaling, high availability, and auto-healing for production workloads, while the break-glass access mechanism provides controlled escalation for critical issues. With SingleStore BYOC, customers can maintain control over their data and cloud infrastructure while benefiting from a fully managed database service.

The MongoDB $avg function is a powerful aggregation operator used to calculate the average value of a field in MongoDB documents. It can be used in various scenarios, including basic queries and advanced aggregations. The operator ignores non-numeric data and handles missing values, making it suitable for calculating averages in different fields. To optimize performance, consider filtering data with a $match stage before performing the aggregation. The $avg function can be combined with other stages like $match, $sort, and $limit to filter, sort, and limit results based on calculated averages. It is commonly used in various use cases such as calculating average grades, sales data, website metrics, sensor data, and financial metrics. Additionally, SingleStore Kai offers faster performance for MongoDB aggregations, including those with $avg, by delivering up to 100x faster performance and scaling horizontally to handle big data and high query volume.

Approximate Nearest Neighbor (ANN) search is a computational technique that quickly finds data points in large datasets similar to a given query point, prioritizing speed and scalability over exact accuracy. ANN is particularly useful for high-dimensional data such as text embeddings, images, and audio, where traditional nearest neighbor searches can be computationally expensive. It works by representing data points as vectors in a multidimensional space, indexing them using specialized algorithms like HNSW or LSH, querying them with similarity metrics, and approximating matches to focus on promising areas of the vector space. ANN search is used in applications such as semantic search, recommendation systems, image search, audio analysis, and fraud detection, offering benefits including scalability, speed, flexibility, cost-effectiveness, and real-time performance. However, it also presents challenges like trade-offs between speed and accuracy, high dimensionality, and indexing overheads. SingleStore integrates ANN search capabilities into its platform, providing a seamless experience for users to try and leverage the technique for faster insights.

Artificial intelligence is evolving beyond automation and predictive analytics to a new paradigm called agentic AI, which enables autonomous, intelligent systems capable of operating with greater autonomy, self-improvement, and adaptability. Agentic AI refers to dynamically refining responses based on real-time data and context, making it valuable for industries requiring rapid adaptability like fintech, cybersecurity, and digital marketing. For businesses, agentic AI represents a significant shift in how AI can be used to drive efficiency and innovation, offering enhanced decision-making, personalization at scale, operational efficiency, and cybersecurity advancements. To fully leverage agentic AI, organizations need a high-performance database infrastructure capable of processing real-time data with ultra-low latency, which is where SingleStore comes in, providing unified storage and compute for AI workloads, low-latency query performance, and scalability to support dynamic AI agents with evolving datasets.

The partnership between Dataloop and SingleStore aims to accelerate the deployment of impactful AI by providing a seamless data lifecycle, from secure data ingestion and preprocessing to model training and evaluation. The integration enables rapid model iteration and deployment with unmatched speed, ensuring consistent and reliable data flow at scale. It supports various AI use cases, including AI agents and multimodal workflows, and facilitates bias analysis and mitigation, synthetic data management, and interactive chatbot development. The modular design of the integration minimizes manual intervention, reduces latency, and ensures models operate on the most current information.

Aura Container Service is a general-purpose compute service that allows users to run containerized applications in a serverless fashion. It powers notebooks and scheduled jobs, providing a flexible and optimized solution for different workloads with varying computation requirements. The service offers several benefits, including instant availability, improved workflow efficiency, GDPR compliance, high availability, and security features such as access control, quota management, and robust security isolation. Aura is designed to support agentic applications, which require foundational Large Language Models, fast databases, general-purpose compute, tools and connectors, and orchestration tools. The platform's future directions include supporting open-source models, server(less?) optimizations, and an integrated AI platform that brings together LLM models, embedding services, and database capabilities.

SingleStore, a data platform designed to build real-time apps and analytics at scale, is revolutionizing genomic data analysis in personalized medicine. The massive datasets generated by genomic research pose significant challenges, including handling vast amounts of data, complex querying, and the need for real-time insights. SingleStore overcomes these challenges with its efficient storage and querying capabilities, fast pattern matching for genetic analysis, and parallel processing of genomic data. By leveraging SingleStore, researchers and healthcare providers can unlock the full potential of genomic data analysis, enabling faster and more accurate insights that lead to improved patient outcomes.

Happy customers are everything to us and we're delighted that SingleStore has been verified as a Trusted Seller on TrustRadius, the most trusted buyer intelligence platform for business technology. The company's commitment to delivering exceptional value, enabling customer success and fostering transparency is recognized by this latest honor following its 2025 Buyer’s Choice Award from TrustRadius. To achieve this status, SingleStore must regularly source reviews, disclose review sourcing methodology and use of incentives, provide equal opportunity for product users to share feedback, read all published reviews and respond when necessary, and regularly update their product profile with up-to-date product information. This recognition comes after SingleStore was awarded the Buyer's Choice Award from TrustRadius, which recognizes companies that are highly rated by customers for capabilities, value for price and customer relationships.