Amazon Bedrock LLM Snap Pack

Overview

The Amazon Bedrock LLM Snap Pack facilitates seamless integration with Amazon's high-performing foundation models (FMs) in the SnapLogic platform. Use this Snap Pack Pack to integrate Amazon Bedrock's FMs from diverse sources with a unified API and leverage Amazon Bedrock's LLM capabilities, such as prompt generation, chat completions, and embedding your SnapLogic workflows.

This Snap Pack includes the following key features:

  • Enables efficient generation of chat completions utilizing Amazon Titan Text and Anthropic Claude models and parameters.
  • Supports tasks like embedding, chat completions, conversational exchanges, and prompt generation for enhanced language processing.
  • Supports the processing and generation of content across multiple modalities, such as text, documents, and images, in a unified workflow.
  • Enables you to query a pre-established Amazon Bedrock knowledge base for relevant context and utilizes models to generate responses.
  • Supports advanced response configurations that enable you to customize response behavior through detailed settings, allowing for precise tuning of outputs based on your specific needs and scenarios.
  • Supports the processing and generation of content across multiple modalities, converting diverse inputs into a standardized response model for seamless integration with LLMs.

This Snap Pack contains the following Snaps:

Snaps Description
Agent Retrieve Retrieves information from a knowledge base based on the specified prompt.
Agent Retrieve and Generate Uses the AWS Bedrock Agent to retrieve information and generate responses from a knowledge base.
Amazon Bedrock Converse API Generates message responses using the specified Amazon Bedrock Converse API model and model parameters.
Amazon Bedrock Converse API Tool Calling Provides a conversational interface with the ability to call external tools or services based on your interaction and context.
Amazon Bedrock Function Generator Generates a tool definition in the output, which needs to be converted into a JSON schema for use with the Amazon Converse API.
Amazon Bedrock Function Result Generator Formats the results generated by user-run functions into a unified structure that must be converted into the specific structure required by a particular provider when calling the tools from the LLM.
Amazon Titan Embedder Generates an embedding vector based on the input data.
Amazon Bedrock Prompt Generator Generates the augmented user prompt as per the specified prompt template.
Amazon Titan Chat Completions Generates chat completions using the specified Amazon Titan Text model and model parameters.
Anthropic Claude on AWS Chat Completions Generates chat completions using the specified Anthropic Claude model and model parameters.
Anthropic Claude on AWS Messages Generates message responses using the specified Anthropic Claude model and model parameters.
Amazon Bedrock Multimodal Content Generator Generates multimodal content for the user prompt with Amazon Bedrock.
Invoke Agent Invokes the Amazon Bedrock Invoke Agent operation on the Amazon Bedrock service.

See it in action