scripts/run_phase2_demo.py

# ============================================================
# FILE: scripts/run_phase2_demo.py
# ============================================================
# PURPOSE:
# Test the modular Phase 2 RAG pipeline from terminal.
#
# RUN:
# python scripts/run_phase2_demo.py
#
# WHAT THIS SCRIPT DOES:
# 1. loads config
# 2. validates config
# 3. creates required folders
# 4. rebuilds vector database
# 5. tests cloud API connection
# 6. asks demo questions
# 7. saves outputs
#
# AI ENGINEER PRODUCTION TIP:
# Always create a simple CLI test script before building the UI.
# This helps confirm backend logic works independently from Streamlit.
# ============================================================

from src.config import create_required_folders, load_config, validate_config
from src.rag_pipeline import RAGPipeline


def print_retrieved_sources(result: dict) -> None:
    """
    Print retrieved source chunks clearly.
    """

    print("\nRetrieved sources:")

    for item in result["retrieved_chunks"]:
        print(
            f"- Rank {item['rank']} | "
            f"Source: {item['source']} | "
            f"Chunk: {item['chunk_index']} | "
            f"Distance: {item['distance']}"
        )


def main() -> None:
    """
    Main demo runner.
    """

    print("=" * 80)
    print("KnowFlow AI - Phase 2 Modular RAG Pipeline Demo")
    print("=" * 80)

    config = load_config()
    validate_config(config)
    create_required_folders(config)

    print("\nConfiguration loaded:")
    print("Project root:", config.project_root)
    print("Data folder:", config.data_folder)
    print("Vector DB folder:", config.vector_db_folder)
    print("Provider:", config.cloud_api_provider)
    print("Model:", config.cloud_chat_model)
    print("Embedding model:", config.embedding_model_name)

    pipeline = RAGPipeline(config=config)

    print("\nRebuilding vector database...")
    rebuild_result = pipeline.rebuild_vector_database()
    print(rebuild_result)

    print("\nTesting cloud LLM connection...")
    connection_answer = pipeline.llm_client.test_connection()
    print("Connection test answer:", connection_answer)

    demo_questions = [
        "What is the refund policy?",
        "How long does standard shipping take?",
        "What does the warranty cover?",
        "Does the company sell customer data?",
        "How many days of annual leave do employees get?",
        "What is the company policy about quantum teleportation?",
    ]

    for question in demo_questions:
        print("\n" + "=" * 80)
        print("Question:")
        print(question)

        result = pipeline.ask(question)

        print("\nAnswer:")
        print(result["answer"])

        print_retrieved_sources(result)

        output_path = pipeline.save_result(result)
        print("\nSaved output:", output_path)

    print("\n" + "=" * 80)
    print("Phase 2 demo completed successfully.")
    print("=" * 80)


if __name__ == "__main__":
    main()