Google Drive

The Google Drive integration lets you browse your Drive, import files into River as entities, and have the AI reason over the contents — all without leaving the workspace.

Two modes#

Browse mode#

Without committing to sync, you can browse, search, and open Drive files inside River. The AI can use Drive tools to fetch a file on demand without permanently importing it. Best for occasional access to a large Drive.

Sync mode#

You pick specific folders to sync continuously. Files in those folders become File entities in your tree, with full text extracted, indexed, and searchable. Best for the folders you actually work in daily.

Setup#

1. Open Settings → Integrations → Google Drive
2. Click Connect Drive and complete OAuth
3. Choose Browse or Sync for each folder
4. For sync, pick the destination Group in your Dataspace

What Drive unlocks for AI#

• "Find the proposal Joey sent last week" — searches contents, not just filenames
• "Summarize the contracts in /Legal/2026" — reads PDFs and Docs, synthesizes
• Auto-extraction — when a contract lands, an Agent extracts terms into a structured Contract entity
• PRD agents — when a new PRD shows up in /Engineering/PRDs, an Agent flags it in #engineering

Supported file types#

River extracts text from:

• Google Docs / Sheets / Slides (native)
• PDFs (text + OCR)
• Word, Excel, PowerPoint
• Markdown, plain text, code files
• Images (with optional OCR)
• Audio and video files (via the Transcripts pipeline)

Anything else is stored as an opaque File entity that you can still open, share, and act on.

Tips#

• For most teams, sync the 2–3 folders you actually work in. Browse the rest.
• Pair with the Browser Extension for clipping web content into the same Drive folders if you want a single source of truth.
• For huge Drives (10k+ files), use Browse mode by default and selectively sync as needed.