Introducing Vector Search Similarity Capabilities in Azure Cache for Redis Enterprise

Created in collaboration with

• Samuel Partee, Principal Applied AI Engineer, Redis
• Kyle Teegarden, Senior Product Manager, Azure Cache for Redis
• Shruti Pathak, Senior Product Manager, Azure Cache for Redis

Introduction 

The latest wave of generative AI, like large language models, has paved the way for significant advancements in the utilization of vector embeddings and vector similarity search. Large language models, such as OpenAI's GPT, can learn complex patterns and representations from vast amounts of text, enabling them to generate rich semantic embeddings for words, sentences, and documents. By leveraging these learned embeddings, developers can harness the power of vector similarity search, revolutionizing how information is organized, retrieved, and analyzed in various domains, including fraud detection, recommendation systems, and information retrieval. 

Today we are thrilled to announce that Azure Cache for Redis Enterprise, now equipped with vector search similarity capabilities, combines the power of a high-performance caching solution with the versatility of a vector database, opening up new frontiers for developers and businesses. 

Azure Cache for Redis Enterprise, a fully managed caching solution, has evolved into more than just a cache—it has transformed into a robust full-purpose database that seamlessly integrates vector search workloads. Now, developers and data professionals can harness the full potential of vector search within Azure Cache for Redis Enterprise, backed by its simplicity, speed, scalability, and reliability. Developers can now use Redis to enable lightning-fast similarity search operations, allowing AI applications to process vast amounts of data and deliver accurate results. 

Understanding Vector Search 

Vector search is a cutting-edge search technique that goes beyond simple keyword matching. Instead of relying on textual queries, vector search utilizes mathematical representations, or vectors, to capture the semantic meaning and relationships between data points. This approach enables fast and accurate similarity search, even in high-dimensional spaces. 

The benefits of vector search are far-reaching. Its speed and scalability make it ideal for real-time applications with large data volumes. For example, in e-commerce, vector search powers personalized product recommendations by quickly identifying similar items based on user preferences. It also plays a crucial role in content matching, fraud detection, and anomaly detection across various industries.  

Leveraging Azure Cache for Redis Enterprise as a Vector Database with OpenAI 

In order to harness the capabilities of vector embeddings and vector similarity search in production environments, the importance of vector databases becomes evident. Vector databases serve as a crucial infrastructure component for efficiently storing, indexing, and querying large volumes of high-dimensional vector data. They leverage advanced indexing techniques, like HNSW,  to enable fast and accurate similarity searches, ensuring efficient retrieval of similar vectors from massive datasets. By integrating vector databases into the production pipeline, organizations can leverage the power of vector similarity search in real-time applications, ranging from recommendation systems and personalized search engines to fraud detection and content analysis. The combination of large language models (LLMs), vector embeddings, and vector databases opens up a new realm of possibilities for leveraging the semantic understanding of textual data and delivering sophisticated applications powered by AI. Developers can use vector databases for contextual retrieval, long-term memory, and semantic caching, all of which are needed to ensure LLM-based applications are accurate, factual, responsive, scalable, and cost-effective.  

As a strategic partner, OpenAI's cutting-edge AI technologies can be seamlessly combined with Azure Cache for Redis Enterprise, enabling you to create intelligent applications that offer personalized recommendations, advanced search capabilities, and natural language understanding. 

Learn more!

More resources regarding Azure Cache for Redis

Get Started Today!

• Azure Cache for Redis Overview    
• Azure Cache for Redis Pricing   
• What’s New in Azure Cache for Redis  

Introductory 

• Intro blog post 
• More In-depth Post on Vector Similarity Search with AI and Azure Cache for Redis
• Solutions page for Redis Vector Database 

VSS Documentation 

• Vector Search API docs 
• Redis-py client Search API docs  

VSS Benchmarks 

• Third-party competitive benchmarks  

VSS Intro Demos 

• Redis Intro in OpenAI Cookbook Repo 

Integrations w/ Redis VSS 

• Jina 
• Langchain 
• LlamaIndex 
• OpenAI Retrieval Plugin 

Use Cases for Redis VSS 

• Visual and Semantic search for e-commerce (hosted link) 
• Semantic Search for Document Retrieval (hosted link) 
• Question and Answer with OpenAI 
• Question and Answer with Azure OpenAI