Google Drive Ingestion

How to pull documents from a Google Drive folder into a client mind, turning brand assets, strategy decks, and meeting notes into searchable knowledge.

Why This Matters#

Clients often share brand guidelines, strategy documents, and historical content via Google Drive. Manually copying and pasting each document into Coppermind works but does not scale when a client shares an entire folder of assets at the start of an engagement.

Drive ingestion lets you point Coppermind at a Google Drive folder and pull everything in at once. Supported files are processed through the same extraction pipeline as transcripts - turning raw documents into structured, searchable memories.

What You Need#

1. A Google account connected to Coppermind. The same Google OAuth connection used for calendar and email integration works here.
2. Access to the Drive folder. You need at least "Viewer" permission on the folder. Shared folders from your client work fine.
3. An active client mind. Switch to the client you want to ingest documents for before starting.

Ingesting a Folder#

Switch to the relevant client, then use the /cm-connect command:

"Pull documents from Google Drive for Acme"

Or specify the folder:

"/cm-connect pull from Google Drive folder 'Acme Brand Assets'"

Coppermind lists the files it finds and classifies each one:

After processing, you get a summary:

Drive ingestion complete for Acme Corp:

Ingested: 5 files
Skipped: 2 files (1 image, 1 oversized PDF)
Errors: 0

Files processed:
- Brand Guidelines v3.docx (12 memories extracted)
- Q1 Strategy Deck.pptx (8 memories extracted)
- Content Calendar.xlsx (4 memories extracted)
- Meeting Notes Jan-Mar.docx (15 memories extracted)
- Competitor Analysis.pdf (6 memories extracted)

When to Use This#

Client Onboarding#

When a new client shares their Google Drive folder with existing brand materials, strategy documents, and historical content:

"Switch to Acme"
"Pull all docs from the Acme shared folder on Drive"

This jumpstarts the client mind with months or years of accumulated knowledge in one step.

Ongoing Document Sync#

When a client adds new documents to a shared folder:

"Check for new documents in Acme's Drive folder"

Previously ingested files are skipped (deduplication by content hash), so only new or updated documents are processed.

Pre-Meeting Prep#

If a client sends a Drive link before a meeting with relevant background:

"Pull the documents from this Drive folder before my meeting with Acme"

Handling Duplicates#

Drive ingestion uses the same deduplication as transcript ingestion:

• Content hash check - If a file has already been ingested (same content), it is skipped
• Semantic dedup - Extracted memories are checked against existing memories to prevent storing near-duplicates

You can safely re-run ingestion on the same folder. Only new or changed files will be processed.

What to Expect#

• Processing time depends on the number and size of files. A folder with 10 documents typically takes 1-2 minutes.
• Memory yield varies by document type. A dense strategy document might produce 15-20 memories. A simple spreadsheet might produce 2-3.
• Not everything is extractable. Images embedded in documents, charts in spreadsheets, and heavily formatted layouts may not produce useful text. The pipeline focuses on readable text content.

Tips#

• Start with the most important folder. You do not need to ingest everything at once. Start with brand guidelines and strategy documents, then expand to other content.
• Check knowledge health after ingestion. Run a knowledge health check to see if the ingested content improved coverage across key topics.
• Ingest before your first real meeting. If a client shares Drive access during onboarding, pull the documents before your first strategy session. Your meeting prep will be dramatically better.
• Tell the client what you found. After ingestion, summarize the key takeaways. This builds trust and shows the value of their document sharing.