# Package index ## Document processing Ingest documents from the web, PDFs, Word, or local Markdown. - [`read_as_markdown()`](https://ragnar.tidyverse.org/reference/read_as_markdown.md) : Convert files to Markdown - [`MarkdownDocument`](https://ragnar.tidyverse.org/reference/MarkdownDocument.md) : Markdown documents - [`ragnar_find_links()`](https://ragnar.tidyverse.org/reference/ragnar_find_links.md) : Find links on a page ## Text chunking Chunk and augment text context. - [`markdown_chunk()`](https://ragnar.tidyverse.org/reference/markdown_chunk.md) : Chunk a Markdown document - [`MarkdownDocumentChunks`](https://ragnar.tidyverse.org/reference/MarkdownDocumentChunks.md) : Markdown documents chunks - [`ragnar_chunks_view()`](https://ragnar.tidyverse.org/reference/ragnar_chunks_view.md) : View chunks with the store inspector ## Store Store embeddings in DuckDB for fast, local queries. - [`ragnar_store_create()`](https://ragnar.tidyverse.org/reference/ragnar_store_create.md) [`ragnar_store_connect()`](https://ragnar.tidyverse.org/reference/ragnar_store_create.md) : Create and connect to a vector store - [`ragnar_store_insert()`](https://ragnar.tidyverse.org/reference/ragnar_store_insert.md) [`ragnar_store_update()`](https://ragnar.tidyverse.org/reference/ragnar_store_insert.md) : Inserts or updates chunks in a `RagnarStore` - [`ragnar_store_ingest()`](https://ragnar.tidyverse.org/reference/ragnar_store_ingest.md) : Concurrently ingest documents into a Ragnar store - [`ragnar_store_build_index()`](https://ragnar.tidyverse.org/reference/ragnar_store_build_index.md) : Build a Ragnar Store index - [`ragnar_store_inspect()`](https://ragnar.tidyverse.org/reference/ragnar_store_inspect.md) : Launch the Ragnar Store Inspector - [`ragnar_store_atlas()`](https://ragnar.tidyverse.org/reference/ragnar_store_atlas.md) : Visualize a store using Embedding Atlas ## Embeddings Embed chunks with commercial or open-source models. - [`embed_azure_openai()`](https://ragnar.tidyverse.org/reference/embed_azure_openai.md) : Uses Azure AI Foundry to create embeddings - [`embed_bedrock()`](https://ragnar.tidyverse.org/reference/embed_bedrock.md) : Embed text using a Bedrock model - [`embed_databricks()`](https://ragnar.tidyverse.org/reference/embed_databricks.md) : Embed text using a Databricks model - [`embed_google_gemini()`](https://ragnar.tidyverse.org/reference/embed_google_vertex.md) [`embed_google_vertex()`](https://ragnar.tidyverse.org/reference/embed_google_vertex.md) : Embed using Google Vertex API platform - [`embed_ollama()`](https://ragnar.tidyverse.org/reference/embed_ollama.md) [`embed_openai()`](https://ragnar.tidyverse.org/reference/embed_ollama.md) [`embed_lm_studio()`](https://ragnar.tidyverse.org/reference/embed_ollama.md) : Embed Text - [`embed_snowflake()`](https://ragnar.tidyverse.org/reference/embed_snowflake.md) : Generate embeddings using Snowflake ## Retrieve Retrieve relevant chunks using both vector and text search. - [`ragnar_retrieve()`](https://ragnar.tidyverse.org/reference/ragnar_retrieve.md) : Retrieve chunks from a `RagnarStore` - [`ragnar_retrieve_bm25()`](https://ragnar.tidyverse.org/reference/ragnar_retrieve_bm25.md) : Retrieves chunks using the BM25 score - [`ragnar_retrieve_vss()`](https://ragnar.tidyverse.org/reference/ragnar_retrieve_vss.md) : Vector Similarity Search Retrieval - [`ragnar_register_tool_retrieve()`](https://ragnar.tidyverse.org/reference/ragnar_register_tool_retrieve.md) : Register a 'retrieve' tool with ellmer - [`mcp_serve_store()`](https://ragnar.tidyverse.org/reference/mcp_serve_store.md) : Serve a Ragnar store over MCP - [`chunks_deoverlap()`](https://ragnar.tidyverse.org/reference/chunks_deoverlap.md) : Merge overlapping chunks in retrieved results