Basic RAG

Create a simple Q&A pipeline with local documents.

This example demonstrates the core functionality of RAGLight: loading documents from a local folder, indexing them, and asking questions using a local LLM (Ollama).

Code Example

1. Setup Configuration

   Import the necessary classes and set up logging to see the pipeline in action.

2. Define Knowledge Base

   Point the pipeline to your data directory. RAGLight handles PDF and text files automatically.

3. Build & Query

   The build() method processes documents and creates the vector store.

main.py

from raglight.rag.simple_rag_api import RAGPipeline
from raglight.config.settings import Settings
from raglight.config.rag_config import RAGConfig
from raglight.config.vector_store_config import VectorStoreConfig
from raglight.models.data_source_model import FolderSource

# 1. Initialize Settings
Settings.setup_logging()

# 2. Configure the vector store (embeddings + ChromaDB)
vector_store_config = VectorStoreConfig(
    embedding_model=Settings.DEFAULT_EMBEDDINGS_MODEL,
    provider=Settings.HUGGINGFACE,
    database=Settings.CHROMA,
    persist_directory="./defaultDb",
    collection_name=Settings.DEFAULT_COLLECTION_NAME,
)

# 3. Configure the RAG pipeline (LLM + retrieval)
config = RAGConfig(
    provider=Settings.OLLAMA,
    llm=Settings.DEFAULT_LLM,
    k=5,
    knowledge_base=[FolderSource(path="./data")],
)

# 4. Build the Pipeline (ingests documents and creates the vector store)
pipeline = RAGPipeline(config, vector_store_config)
pipeline.build()

# 5. Ask a Question
response = pipeline.generate("What are the key takeaways from these documents?")
print("\n🤖 Answer:", response)