Skip to main content

​
Using RAG in Rowboat

Rowboat provides multiple ways to enhance your agents’ context with Retrieval-Augmented Generation (RAG). This guide will help you set up and use each RAG feature.
RAG is called “Data” on the build view in the Rowboat UI.

​
Types of RAG

RAG TypeDescriptionConfiguration Required
Text RAGProcess and reason over text content directlyNo configuration needed
File UploadsUpload PDF files directly from your deviceNo configuration needed
URL ScrapingScrape content from web URLs using FirecrawlRequires API key setup
URL Scraping does not require any setup in the managed version of Rowboat.
Adding data sources for RAG in Rowboat

​
RAG Features

​
1. Text RAG

Process and reason over text content directly

​
2. File Uploads

  • Upload PDF files directly from your device
  • Open Source Version: Files are stored locally on your machine
  • Managed Version: Files are stored in cloud S3 storage
  • Files are parsed using OpenAI by default
You can also use Google’s Gemini model for parsing as it is better at parsing larger files.

​
2.1 Using Gemini for File Parsing

To use Google’s Gemini model for parsing uploaded PDFs, set the following variables: 
# Enable Gemini for file parsing
export USE_GEMINI_FILE_PARSING=true
export GOOGLE_API_KEY=your_google_api_key

​
3. URL Scraping

Rowboat uses Firecrawl for URL scraping. You can have a maximum of 100 URLs. Open Source Version: To enable URL scraping, set the following variables: 
export USE_RAG_SCRAPING=true
export FIRECRAWL_API_KEY=your_firecrawl_api_key
Managed Version: No configuration required - URL scraping is handled automatically.