ContextGem

ContextGem is a robust framework designed to streamline the process of extracting structured data and insights from documents using Large Language Models (LLMs). It addresses the challenges faced in traditional LLM frameworks, which often require extensive boilerplate code, leading to increased development time and complexity.

Why Choose ContextGem?

ContextGem simplifies the extraction of structured information by offering flexible and intuitive abstractions. This advanced framework minimizes development overhead by eliminating boilerplate code and efficiently managing the complex aspects of data extraction. Users can leverage built-in features like automated dynamic prompts, precise reference mapping, and justifications based on extraction reasoning.

Key Features

• Built-in Abstractions:
  • Automated dynamic prompts
  • Automated data modeling and validators
  • Precise granular reference mapping (paragraphs & sentences)
  • Justifications for reasoning behind extractions
  • Neural segmentation (SaT)
  • Multilingual support without prompting
  • Unified and fully serializable extraction pipeline
  • Support for multiple LLM providers

See descriptions of ContextGem abstractions and comparisons of specific implementation examples using ContextGem and other popular open-source LLM frameworks.

Quick Start Example

A concise example of utilizing ContextGem for extracting anomalies from a document is as follows:

import os

from contextgem import Document, DocumentLLM, StringConcept

# Example document instance
# Document content is shortened for brevity
doc = Document(
    raw_text="""
    Consultancy Agreement\n
    This agreement between Company A (Supplier) and Company B (Customer)...\n
    The term of the agreement is 1 year from the Effective Date...\n
    ...\n    The purple elephant danced gracefully on the moon while eating ice cream.\n    ...\n    """
)

# Attach a document-level concept

doc.concepts = [
    StringConcept(
        name="Anomalies",  # Anomalies in the document
        description="Anomalies in the document",
        add_references=True,
        reference_depth="sentences",
        add_justifications=True,
        justification_depth="brief",
    )
]

# Create an LLM for data extraction
llm = DocumentLLM(
    model="openai/gpt-4o-mini",  # Specify the LLM model
    api_key=os.environ.get("CONTEXTGEM_OPENAI_API_KEY"),  # API key for the LLM provider
)

# Extract information from the document
doc = llm.extract_all(doc)

# Access extracted items
print(doc.concepts[0].extracted_items)  # Extracted items with references and justifications

For more examples and comprehensive usage documentation, visit contextgem.dev.

Focused Document Analysis

ContextGem excels in detailed document analysis by leveraging LLMs' extended context capabilities. This approach maximizes extraction accuracy for individual documents, capitalizing on evolving LLM technology to minimize retrieval inconsistencies and optimize for thorough, single-document evaluations. It is important to note that ContextGem currently does not support cross-document querying.

Optimizations

To enhance performance and reduce costs, the ContextGem documentation provides guidelines for optimization strategies across various dimensions:

• Optimizing for Accuracy
• Optimizing for Speed
• Optimizing for Cost