def retrieve_info()

in Runtime_env/app/orchestration/tools.py [0:0]

• 5 lines of code
• 3 McCabe index (conditional complexity)

def retrieve_info(query: str) -> tuple[str, List[Document]]:
    """
    Use this when you need additional information to answer a question.
    Useful for retrieving relevant information based on a query.

    Available documents:
       Finance: `Alphabet Investor PDFs`: This data contains quarterly earnings releases
        and annual reports for Alphabet for every quarter since 2004. The Annual reports include 
        financial statements (balance sheet, income statement, cash flow statement), a letter to
        shareholders, management discussion and analysis (MD&A), and information on corporate governance.
        The quarterly earnings releases also contain key financial statements like the income statement,
        balance sheet, and cash flow statement, along with management commentary and analysis of
        quarterly performance.
        Dataset can be found here:
        https://console.cloud.google.com/storage/browser/cloud-samples-data/gen-app-builder/search/alphabet-investor-pdfs

       HealthCare: `PriMock57 Healthcare consultations`: This dataset consists of 57
        mock medical primary care consultations held over 5 days by 7
        Babylon clinicians and 57 Babylon employees acting as patients,
        using case cards with presenting complaints, symptoms, medical
        & general history etc.
        Dataset can be found here:
        https://console.cloud.google.com/storage/browser/cloud-samples-data/vertex-ai/medlm/primock57/transcripts

      Retail: `Google Store`: This data is a list of html web pages from the Google Store from
        2023. It represents a listing of products, details, prices, etc related to
        Google products.
        Dataset can be found here:
        https://console.cloud.google.com/storage/browser/cloud-samples-data/dialogflow-cx/google-store

    Args:
        query (str): The user's question or search query.

    Returns:
        List[Document]: A list of the top-ranked Document objects, limited to TOP_K (5) results.
    """
    # Use the retriever to fetch relevant documents based on the query
    retrieved_docs = lang_retriever.invoke(query)
    # Re-rank docs with Vertex AI Rank for better relevance
    ranked_docs = compressor.compress_documents(documents=retrieved_docs, query=query)
    # Format ranked documents into a consistent structure for LLM consumption
    formatted_docs = format_docs.format(docs=ranked_docs)
    return (formatted_docs, ranked_docs)