Using Converse API¶

The Converse API provides streaming capabilities for real-time interactions with documents. This guide demonstrates how to use the Converse API with Rhubarb.

Basic Usage¶

Here’s a simple example of how to use the Converse API:

 from rhubarb import DocAnalysis, SystemPrompts
 import boto3

 # Initialize a boto3 session
 session = boto3.Session()

 # Create a DocAnalysis instance with converse API
 da = DocAnalysis(
     file_path="./test_docs/employee_enrollment.pdf",
     boto3_session=session,
     use_converse_api=True,
     system_prompt=SystemPrompts().SummarySysPrompt
 )

 # Stream the response
 for resp in da.run_stream(message="Give me a brief summary of this document."):
     if isinstance(resp, str):
         print(resp, end='')
     else:
         print("\n")
         print(resp)

Key Components¶

DocAnalysis Parameters¶

file_path: Path to the document you want to analyze
boto3_session: AWS boto3 session for API authentication
use_converse_api: Boolean flag to enable streaming responses (default: False)
system_prompt: System prompt to guide the analysis

Streaming Responses¶

The run_stream() method yields responses in real-time. The response can be either:

A string containing part of the generated text
A dictionary containing metadata or completion information

Configuration Options¶

System Prompts¶

You can customize the behavior using different system prompts:

# Using custom system prompt
custom_prompt = "Analyze this document and provide key insights."
da = DocAnalysis(
    file_path="document.pdf",
    boto3_session=session,
    use_converse_api=True,
    system_prompt=custom_prompt
)

# Using built-in system prompts
da = DocAnalysis(
    file_path="document.pdf",
    boto3_session=session,
    use_converse_api=True,
    system_prompt=SystemPrompts().SummarySysPrompt
)