Redis for AI Background

Redis was built to address the need for speed and scale in the emerging mobile and cloud era.Its founder, Salvatore Sanfilippo, began developing Redis in 2009, initially as an internal tool for scaling his own startup. Like many side projects, Redis quickly became a star product.

The problems Salvatore was trying to solve at the time have many similarities to the challenges organizations face today as they build modern AI technology stacks. redis helped meet the new demands for speed and scale in the mobile and cloud era. Now we're working to do the same for GenAI. We are at the beginning of the second act of Redis - a new era of high speed, where Redis becomes an important part of the new technology stack for AI.

Redis for AI: Products, Tools, and Support for Building Rapid GenAI Applications

Figure 2 - Redis for AI

Bringing GenAI applications into production and scaling them is a daunting task. To simplify the process, we've introduced the Redis for AIThis is an integrated suite. This is an integrated suite that includes all of the AI capabilities of Redis, as well as the support needed to bring GenAI applications into production.

Developers can use Redis for AI to:

• RAG with the World's Fastest Vector Database: Building a Real-Time Architecture for GenAI Applications.
• Semantic Caching: Quickly retrieve saved answers, saving over 30% of costly LLM calls.
• LLM Memory: Personalize user sessions to provide all the data needed for LLM in a timely manner.
• Agent Memory: Allows agents to perform complex reasoning faster and provide better answers.
• Feature storage: Get predictions from production-grade ML models in less than 1 millisecond.

Redis for AI includes dependencies, reference architectures, the new RedisVL 0.3.0, and new specialized partner packages such as langchain-redis and llama-index-vector-stores-redis. You'll also benefit from our robust partner ecosystems, which are built on our real-time data platform and benchmark-leading vector databases.

Redis for AI has the same flexibility as Redis, supports a wide range of data types, and runs in cloud, local, and hybrid environments. Our specialized support allows you to work with experts who have implemented Redis for AI in large companies around the world.

Experts say: We can't scale ChatGPT without Redis.

Figure 3 - AI Technology Interface

Increase cache capacity by 5x, from 1 GB to TB, for the same price.

In the past, Redis Auto Tiering has only been used for large caches. Developers want to use Redis in more areas, but are forced to keep caches small due to cost. Smaller caches often require special cache policies and application logic to handle the limitations and result in more cache misses. Now, with Redis Cloud Essentials, you can get 5 GB of cache for the price of 1 GB.

With the acquisition of Speedb this year, Redis was able to reduce the cost of caching at all scales. We've been analyzing technology trends and have found that major advances in solid state drives (SSDs) can bring even more value to Redis users.Redis Flex runs on DRAM and SSDs and is faster and cheaper than other memory solutions such as ElastiCache, Memorystore and even Community Edition.So whether you use Redis Software or Redis Cloud, you can reduce your costs by up to 80% compared to self-managed Redis without changing your existing data infrastructure.

Redis 8: 15 Years of Fast, Flexible, Free Infrastructure

Millions of developers around the world use our Open Source Community Edition (CE). It's the foundation on which we build everything, but many developers say it can be a little difficult to get started. There are too many options, too many client-side complexities, and a lack of clear best-practice guidelines. Recently, there have been a few new branches of Redis, but our community edition is still state-of-the-art, with tons of new features and good support. Now it's faster and easier to get started with Redis.

Figure 4 - Redis Data Flow Visualization

Redis Stack has become the preferred way for users to get started with Redis, and now all of its features have been integrated into Redis Community Edition. When you start using Redis software or the Redis Cloud, you'll get a streamlined, efficient experience (Redis 8 will be released this fall).

Redis 8 will help you get up and running faster and includes all the features of Redis 7.4 CE and Redis Stack 7.4:

• Store, update, and retrieve JSON data:Directly query JSON data for fast, flexible operations and dynamic interactions.
• Build smarter, faster GenAI applications with vector databases:Use the world's fastest vector database. The enhanced Redis query engine increases past query throughput by a factor of 16, making Redis faster than any other vector database in our benchmarks.

Figure 5 - RedisAI Integration Icon

• Search and query data:Quickly find, filter, summarize and sort through millions of records, turning data into actionable insights.
• Optimize time series data:Efficiently collect, store and analyze time series data with high throughput and minimal latency for real-time insights and analysis.
• Handles data streams and large data sets:Fast approximation of counts, frequencies, and rankings using Redis' probabilistic data structure.

Redis Copilot: Your new Redis AI assistant to accelerate the development process

Developers need to build applications faster than ever before. To improve efficiency, we have developed Redis CopilotCopilot is free and now available in Redis Insight, your virtual assistant for all things Redis. We've also released extensions to Visual Studio Code that support Redis, making it easy to build with Redis in an integrated development environment (IDE).

Redis Copilot can help:

• Retrieve information from documents faster:Get best-practice answers on how to format a query or set up a cluster.
• Automatic code writing:Tell Copilot what you need and it will generate code snippets or commands to fulfill your requirements.
• Query data:Make a request in natural language and Copilot will write a query to answer your question.

Like many of our customers who are building their own virtual assistants, we built Redis Copilot to provide fast response in large-scale environments.Redis Copilot leverages the Redis Instant Data Platform and Redis for AI's semantic caching, vector database, and rate-limiting capabilities to provide you with fast, accurate responses.

Redis Copilot is now available, contact us to get it now!

Redis for VS Code will be available soon, look forward to it.

Redis Data Integration: Redis Cloud to be released, accelerating existing databases with automated data pipelines

We recently released Redis Data Integration (RDI)It accelerates your applications and provides data integration without having to redesign your overall data strategy.RDI is now available in private preview on the Redis Cloud and officially released on Redis software.

RDI uses automated data pipelines to synchronize data from existing databases with Redis with minimal setup. It provides connectivity through a single application interface, allowing you to streamline the development process and improve data reliability. We've done the heavy lifting, so you can make your applications run faster without investing a lot of time, effort, and money into building your own data pipeline.

Build applications quickly with our latest innovations!

Figure 6 - AI Microprocessor

We are very excited about the second act of Redis. We're enabling you to build GenAI applications while leveraging AI in our products, it's a new era of speed and Redis is an important part of the new technology stack for AI.

In addition, to keep you up-to-date with the latest innovations faster, we'll be releasing smaller versions more frequently, so stay tuned. Most importantly, we won't sacrifice the past to build the future. Whether you're using Redis to power a high-speed AI application or as a high-availability database or cache, we'll help you build applications fast.