Power generative AI applications with simple, scalable, and high-performing vector store and similarity search
The Vector Engine for Amazon OpenSearch Serverless introduces a simple, scalable, and high-performing vector storage and search capability that helps developers build machine learning (ML)–augmented search experiences and generative artificial intelligence (AI) applications without having to manage the vector database infrastructure. Get contextually relevant responses across billions of vectors in milliseconds by querying vector embeddings, which can be combined with text-based keywords in a single hybrid request.
With a fully-managed Retreival Augmented Generation (RAG) offered by Knowledge Bases for Amazon Bedrock, you can securely connect foundation models (FMs) to your company data stored as embeddings in the vector engine for more relevant, context-specific, and accurate responses without continuously retraining the FM.
Benefits
Increase developer productivity
Colocate your vector and text search to easily query embeddings, metadata, and descriptive text within a single call, increasing search accuracy and reducing system complexity.
Store vector embeddings trained on your business data
Elevate your customer experiences with highly relevant and accurate responses generated by search results based on vector embeddings trained on your business data.
Update vector embeddings in a production-ready search application
Add, update, and delete vector embeddings in near real time without impacting query performance or re-indexing data.
Enhance scalability and efficiency
Store and search billions of vector embeddings with thousands of dimensions in milliseconds with the simplicity of a highly performant and easy-to-use serverless environment.
Use cases
Recommendation engines
Anticipate your customers' needs and provide personalized search experiences geared toward their interests, powered by RAG.
Visual search
Improve your decision-making process with object recognition and analysis.
Personalized chatbots
Provide interactive responses and assistance to better support your customers, powered by RAG.
Fraud detection
Identify and prevent suspicious activity or transactions through efficient comparisons of similar past fraudulent events.
Customers and partners
-
riskCanvas
-
Academia
-
Intuit
-
riskCanvas
-
riskCanvas is a subsidiary of Genpact. It is a SaaS product offering for a financial crime compliance solution that uses cutting-edge big data, automation, and machine learning technologies to deliver compliance, efficiency, and automation to its clients.
"riskCanvas directly integrates with the vector engine for Amazon OpenSearch Serverless, allowing us to expose our existing client operational data through AWS’s generative AI capabilities. This is a game changer, as we can now leverage summarization to accelerate the analysis of investigations, author seed narratives of financial crimes reporting, and make recommendations on escalations—all while using real data kept contained within the riskCanvas secure enclave. With the vector engine, we are reducing handle time across financial crimes use cases, improving consistency of narratives with fewer errors, driving higher efficacy via straight-through processing, and shifting human involvement to deeper analysis."
Ryan Skousen, Chief Technology Officer (riskCanvas) and Vice President of Technology, Genpact Financial Crimes
-
Academia
-
Academia is a platform for sharing academic research. The mission of Adademia.edu is to accelerate the world's research.
“Amazon OpenSearch Service advances Academia’s mission of accelerating the world’s research by enabling us to efficiently index and search through millions of vectors and find the most relevant academic papers to recommend to our users. Switching to Amazon OpenSearch Service has driven a 20% increase in engagement from our users with our content recommendations compared to our previous recommendation solutions. "
Bob Tucker, Director of Engineering, Academia
-
Intuit
-
Intuit Inc. is a global financial technology platform that helps consumers and small businesses prosper by delivering financial management, compliance, and marketing products and services.
"Our platform team worked closely with AWS to build high-level capabilities to efficiently store, manage, and query vector embeddings produced by state-of-the-art ML models, unlocking new possibilities for natural language processing applications and services. This solution is now the default store for all vector needs across Intuit - thanks to Amazon OpenSearch Service. We are excited to expand our OpenSearch-based vector database adoption to tackle new and upcoming use cases in the coming months."
Achal Kumar, Director of Data Capabilities, Intuit
Resources
The role of vector data stores in generative AI applications
What is a vector database?
Fully managed RAG experience with Amazon OpenSearch Serverless and Amazon Bedrock
Amazon OpenSearch Service vector database capabilities explained
Amazon OpenSearch Service for vector search
Using OpenSearch as a vector database
Product search demo for OpenSearch