Amazon Bedrock - RAG and knowledgebase Flashcards
(43 cards)
What does RAG stand for?
Retrieval Augmented Generation
RAG allows foundation models to reference external data sources without fine-tuning.
What is the role of Amazon Bedrock in RAG?
It builds and manages the knowledge base
Amazon Bedrock automates the creation of knowledge bases using data sources like Amazon S3.
What type of database backs the knowledge base in RAG?
Vector database
The vector database allows for retrieval of relevant information from the knowledge base.
What is the process involved in creating data for the Vector database?
Creating Vector embeddings
Bedrock automates the process of generating vector embeddings from data.
What types of data sources can be used with Amazon Bedrock?
- Amazon S3
- Confluence
- Microsoft SharePoint
- Salesforce
- Webpages
These sources can include various types of files and web content.
What happens when a user asks a question to the foundation model?
The model searches the knowledge base for relevant information
This process is done automatically behind the scenes.
What is an example of a specific query a user might ask?
Who is John’s product manager?
Im assuming that stefane is refering to a particular manager to whome a
This query illustrates a specific and contextual request for information.
What are two services on AWS that can be used as Vector databases?
- Open Search Service
- Amazon Aurora
These services are designed to handle vector data efficiently.
Name three other options for Vector databases.
- MongoDB
- Redis
- Pinecone
These options offer additional flexibility for database management.
What is the function of the embeddings model in RAG?
To convert data into vectors
The embeddings model can differ from the foundation model.
What is the preferred choice for high performance and real-time similarity queries?
OpenSearch Service
It offers scalable index management and fast nearest neighbor search capabilities.
What are the two relational databases mentioned?
- Amazon Aurora
- Amazon RDS for PostgreSQL
Both are suitable for relational database needs.
What is a key feature of Amazon DocumentDB?
MongoDB compatibility
It provides real-time similarity queries and can store millions of vector embeddings.
What type of database is Neptune?
Graph database
Neptune is specifically designed for graph data management.
Fill in the blank: Amazon Bedrock can be used to build a _______ that answers customer queries.
ChatBot
The ChatBot utilizes a knowledge base to provide responses.
What is a use case for RAG in legal research?
Providing information on laws, regulations, and case precedents
RAG can facilitate legal queries through a specialized ChatBot.
What kind of data can be included in a healthcare knowledge base?
- Diseases
- Treatments
- Clinical guidelines
- Research papers
- Previous patient data
This data helps answer complex medical queries.
True or False: RAG can be used to answer complex medical queries.
True
RAG applications can be designed for healthcare-related inquiries.
What is AWS Bedrock?
AWS Bedrock is a fully managed service that allows users to build and scale generative AI applications using foundation models.
True or False: AWS Bedrock supports only one type of foundation model.
False
Fill in the blank: AWS Bedrock provides access to a variety of __________ models.
foundation
What is RAG in the context of AWS Bedrock?
RAG stands for Retrieval-Augmented Generation, a technique that combines retrieval and generation for improved AI performance.
Which AWS service is primarily used for data storage when implementing RAG?
Amazon S3
Multiple Choice: Which of the following is a benefit of using AWS Bedrock?
A) Automatic scaling
B) Limited model access
C) High latency
D) Complex integration
A) Automatic scaling
