> ## Documentation Index
> Fetch the complete documentation index at: https://docs.ragrails.com/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Search Engine

> Semantic search that understands meaning, not just keywords.

## The problem

Keyword search misses results that use different words for the same idea. "How do I get my money back" may not match a page titled "Refund policy."

## Why Ragrails fits

Ragrails can run retrieval without answer generation. That means you can build semantic search with no LLM generation cost, then add reranking when result order matters.

| Need                    | Ragrails capability                                                   |
| ----------------------- | --------------------------------------------------------------------- |
| Meaning-based search    | [Embedding](/features/embedding) and [Retrieval](/features/retrieval) |
| Better ordering         | [Reranking](/capabilities/reranking)                                  |
| No generated answer     | Use `retrieve()` or `query()` instead of `chat()`                     |
| Structured result cards | Use `items[].text`, `items[].metadata`, `score`, and `rerank_score`   |

## Sample

<CodeGroup>
  ```python SDK theme={null}
  result = rag.retrieve(
      "how do I get my money back",
      top_k=20,
      use_rerank=True,
      rerank_top_k=5,
  )

  for hit in result.items:
      print(hit.rerank_score, hit.metadata.get("title"), hit.text[:120])
  ```

  ```bash CLI theme={null}
  ragrails retrieve "how do I get my money back" \
    --vector-db qdrant \
    --collection docs \
    --url http://localhost:6333 \
    --provider voyage \
    --model voyage-3 \
    --top-k 20 \
    --rerank \
    --rerank-top-k 5
  ```

  ```bash REST API theme={null}
  curl -X POST http://127.0.0.1:8000/v1/retrieve \
    -H "Content-Type: application/json" \
    -d '{
      "query": "how do I get my money back",
      "provider": "voyage",
      "model": "voyage-3",
      "vector_db": "qdrant",
      "collection": "docs",
      "url": "http://localhost:6333",
      "top_k": 20,
      "use_rerank": true,
      "rerank_top_k": 5
    }'
  ```
</CodeGroup>

## Result card shape

```json theme={null}
{
  "title": "Refund policy",
  "snippet": "Refunds are available within 30 days...",
  "score": 0.76,
  "rerank_score": 0.93,
  "source": "policy/refunds.md"
}
```

<Tip>Add [Chat](/features/chat) later when you want generated answers over the same index.</Tip>