Quickstart

uv add lexigram-ai-rag

Optional extras for PDF loading, web scraping, reranking, and compression:

uv add "lexigram-ai-rag[pdf,web,compression,reranking]"

---

Minimal Example

import asyncio
from lexigram import Application
from lexigram.ai.rag import RAGModule


async def main() -> None:
    async with Application.boot(
        name="rag-demo",
        modules=[RAGModule.configure()],
    ) as app:
        pipeline = await app.container.resolve(RAGPipelineProtocol)
        result = await pipeline.execute(
            RAGContext(query="What is Lexigram?")
        )
        if result.is_ok():
            print(result.unwrap().answer)


asyncio.run(main())

---

What Just Happened

1. RAGModule.configure() created a RAGProvider with default config (in-memory backing, top_k=5, chunking enabled).
2. The provider registered RAGPipelineProtocol and RetrievalStrategyProtocol in the container.
3. pipeline.execute() ran retrieval → synthesis → quality assurance and returned a RAGResponse.

---

Next Steps

• Guide — mental model, core concepts, common patterns
• Configuration — all config keys and env vars
• How-Tos — copy-pasteable recipes