May 2025 Summaries

The latest Redis releases include significant performance improvements, with version 8 being the fastest yet, offering over 30 enhancements. The general availability of Azure Managed Redis (AMR) marks a major milestone, providing a fully-managed, first-party Redis offering co-engineered in partnership with Microsoft Azure. Additionally, updates have been made to Redis Cloud and Private Service Connect, enhancing customer control and performance. For those interested in learning more about RDI, a new lab has been released, aiming to make users an expert in the tool within an hour.

Financial institutions are under pressure to modernize due to rising customer expectations, evolving threats, and growing data volumes. Over 60% of senior banking decision-makers admit that their slow digital transformation has resulted in missing out on winning new customers. To address this, leading banks and fintechs turn to Redis for speed, scalability, and intelligence. Mobile-first users demand real-time app experiences, while AI-driven customer services are becoming table stakes. Financial institutions use Redis as a session store or cache alongside existing systems to improve performance and scalability without rearchitecting. This helps with mobile banking, fraud detection, and customer service by providing instant access to data, detecting fraud instantly, and powering AI-driven innovations. With Redis's sub-millisecond latency, decisions are made instantly, saving millions in potential losses each year. The technology also enables semantic caching, improving response accuracy and performance, particularly for AI-powered applications like chatbots. By leveraging Redis, financial institutions can deliver real-time experiences, fight fraud instantly, and scale AI-driven innovations without rearchitecting from scratch.

The Model Context Protocol (MCP) is a standard developed by Anthropic that enables AI agentic apps to access and work with external data and tools, effectively acting as a universal adapter. MCP integrations are available on various platforms such as GitHub Copilot, Cursor, Augment, and the OpenAI Agents SDK, allowing developers to tap into real-time information like weather forecasts or live stock prices. The Redis' MCP servers have been introduced through two open-source projects: mcp-redis and mcp-redis-cloud, which provide natural language interfaces to manage and search data in Redis. These servers enable AI-driven workflows to interact with structured and unstructured data, making it easier for developers to build powerful agentic apps that are context-aware and capable of managing complex tasks.

Redis excels in GenAI applications by delivering real-time performance and low-latency lookups, making it an ideal choice for applications that require immediate updates to vector data and metadata. Unlike Elasticsearch, Redis applies changes directly to data structures in memory, eliminating the need for manual shard management, reindexing, and JVM tuning. This simplifies operational effort and reduces infrastructure costs, resulting in a better user experience. With its built-in features like TTL support, caching, and automatic sharding, Redis provides a scalable and high-throughput platform that can handle large volumes of data and frequent updates, making it the superior choice over Elasticsearch for GenAI applications.

This post showcases how to use Redis in Spring AI 1.0, a comprehensive solution for AI engineering in Java, to build fast and efficient AI apps that scale. The integration of Redis with Spring AI enables developers to leverage the power of vector storage, retrieval augmented generation, and Model Context Protocol (MCP) support. With this setup, developers can create AI-powered applications that can process large amounts of data efficiently, making it an ideal solution for various use cases such as chatbots, content moderation, and more.

Azure Managed Redis is a new fully-managed, first-party Redis offering co-engineered with Microsoft Azure, allowing customers to access the latest features and advancements while ensuring full compatibility with previous versions. It's built on Redis 7.4 and supports both traditional caching and AI-specific workloads, featuring vector data structures, exact matching, geospatial queries, and fast data processing. The service offers a seamless integration with Microsoft Azure ecosystems, including Azure AI, and provides access to client libraries for popular programming languages, as well as tools like Redis Insight and Redis Copilot to help developers build and optimize their Redis applications. With its advanced features and compatibility, Azure Managed Redis is designed to deliver better performance and scalability at a lower cost than previous Redis services on Azure.

In the last 15 years of its evolution, Redis has adapted to and helped shape three major computing paradigms: client-server architecture, distributed systems, and serverless computing. Initially created as a caching layer in 2009, Redis evolved into a comprehensive data platform that addressed key challenges in each era. From solving client-server bottlenecks to enabling AI apps, Redis consistently offered elegant solutions to pressing needs of each technological age. Its evolution demonstrates how a simple idea can become a powerful tool for developers worldwide, making the impossible routine and improving development experiences.

Red Hat OpenShift AI and Redis offer a powerful combination for organizations looking to use artificial intelligence effectively. By providing a flexible gen AI development environment and a high-performance data platform, these technologies empower data scientists and machine learning engineers to build innovative AI solutions that create real business value. This is achieved through the deployment of Redis as a vector database and semantic cache, enabling efficient retrieval of LLM responses based on the meaning of the query, reducing costs and improving scalability for AI-powered applications. Additionally, OpenShift AI provides an environment for data scientists to use different tools, including embedding models, third-party frameworks like LangChain or LlamaIndex, and multiple LLMs to implement gen AI use cases at scale, while Redis delivers sub-second latency that gen AI use cases need. This enables the creation of more engaging and context-aware chatbots, personalized customer experiences, and improved ability to handle complex conversations. The combination of OpenShift AI and Redis also simplifies deployment and management of a comprehensive set of AI/ML tools, providing organizations with an efficient way to deploy and manage their AI infrastructure.

The rise of hyperscalers has created a challenge for open source companies like MongoDB and Elastic, as they struggle to maintain innovation and investment in their OSS projects while cloud providers reap profits without proportional contributions. To address this, Redis initially took a separate approach with its Redis Stack distribution under a different license, but ultimately decided to adopt the SSPL license to enhance its core capabilities. This change hurt the Redis community's relationship with the company, but has led to new collaborations and improvements, including the rejoining of Salvatore Sanfillipo as a developer evangelist, and the adoption of AGPL licensing for Redis 8, demonstrating a commitment to advancing the platform according to its original vision.

The latest updates from Redis include a new Spring Release with tools and frameworks for Redis for AI, major updates to Redis Cloud, and the release candidate of Redis 8. A fully managed semantic cache is now in private preview, allowing developers to speed up their GenAI apps by reusing past responses via a simple REST API. Additionally, Redis introduces a native data type that treats each element as a vector, making similarity search easier, and integrates directly into LangGraph's memory architecture for improved performance. The company has also expanded its offerings in the cloud with new releases of Redis Cloud Pro and Redis Insight on cloud, providing enterprise-grade CDC-powered caching and centralized observability. Furthermore, developers can now run Redis Cloud in their own VPC on AWS for full control over networking and data locality.

Redis 8, the latest version of Redis, is now generally available in Redis Open Source. It boasts over 30 performance improvements, including up to 87% faster commands, 2x more operations per second throughput, and up to 16x more query processing power with horizontal and vertical scaling. The new version introduces several new data structures such as vector set, JSON, time series, and probabilistic structures, which enhance the capabilities of Redis for AI use cases. Additionally, Redis 8 combines its Redis Stack and community offerings into a single Redis Open Source distribution, providing a unified feature set to users everywhere. This release also includes improvements in replication, multithreading, and I/O threading, making it more efficient and scalable than previous versions. The new version is available for free under the AGPLv3 license and can be installed through various channels as they become available.