Google Gemini_ How to Connect and Collect

Google Gemini Overview

Onna's Google Gemini connector allows you to collect and preserve AI conversation data generated by Google Gemini within your organization's Google Workspace environment. Because Google does not provide a direct API for Gemini content, Onna retrieves this data through the Google Vault eDiscovery API — the same compliance infrastructure used to preserve and export Google Workspace content.

Once collected, Gemini conversations become searchable within Onna, giving your team visibility into AI-assisted interactions across your organization for eDiscovery, compliance, and investigation purposes.

Connector Features
Authorized Connection Required? Yes	Is identity mapping supported? No
Audit logs available? Yes	Admin Access? Yes
Supports a full archive? Yes	Custodian based collections? Yes
Sync modes supported: One-time sync, Auto-sync, Auto-sync and archive	Is file versioning supported? No

Google Gemini Requirements

To connect and collect Google Gemini data in Onna, the following must be in place:

Google Workspace account with Google Vault enabled
Google Gemini for Workspace enabled for your organization
A Google Vault admin account with permissions to create and manage exports
An authorized connection configured in Onna using a service account with the appropriate Google Vault scopes

Note
Since Gemini content is retrieved through Google Vault, your organization's Vault retention rules and hold policies apply to what can be collected. Content not covered by a Vault retention policy may not be available for export.

How to Connect and Collect Onna using Google Gemini

Google Vault is asynchronous by nature, the Gemini collection process runs differently from connectors that pull data directly from an API. Here is what happens during a sync:

Step 1 — Custodian resolution Onna identifies the custodians included in your sync and prepares export requests for each one.

Step 2 — Export creation via Google Vault Onna submits an export request to the Google Vault eDiscovery API for each custodian. Google then processes the export in the background, which can take anywhere from a few minutes to several hours depending on the volume of data.

Step 3 — Export monitoring Onna monitors the status of each export automatically. No action is required on your part while exports are processing. This state is managed by Onna's workflow infrastructure, which means progress is preserved even if the sync is interrupted.

Step 4 — Download and ingestion Once Google signals that an export is ready, Onna downloads the resulting files and parses the conversation data contained within them.

Step 5 — Search and access Ingested conversations become searchable within Onna alongside your other collected content. They can also be exported in a consistent format compatible with other content types in Onna.

Note
Exports are processed by Google Vault and not fetched in real time, a delay between a conversation occurring and it appearing in Onna is expected. The length of this delay depends on export processing times on Google's side.

What Gets Collected

Onna collects Google Gemini AI conversations as dedicated conversation resources. Each conversation thread is ingested as a standalone item that can be searched, reviewed, and exported independently.

Content	Details
Conversation threads	The full exchange between a user and Gemini, including prompts and responses, preserved in chronological order
Participants	The Workspace user (custodian) associated with the conversation
Timestamps	Creation and modification dates for each conversation
Model information	Metadata identifying the Gemini model involved
Folder path	Structured path used to organize conversations by custodian within the collection

Note
User and AI interactions are clearly distinguishable within each collected thread, preserving the conversational structure for review.

Google Gemini Considerations & Limitations

Export Processing Time

Google Vault exports are not instant. Depending on the volume of Gemini conversations for a given custodian, exports can take minutes to several hours to complete. Onna waits for exports to finish before ingesting content, so large collections may take longer than other connector types.

Vault Retention Policies Apply

Only conversations covered by your organization's Google Vault retention rules or active legal holds are available for collection. Conversations outside the scope of a Vault policy may not be retrievable.

Content Availability Window

Gemini conversations are only available through Vault if they fall within the retention period defined in your Vault configuration. Conversations older than the retention window will not be returned in exports.

No Real-Time Collection

Because all data retrieval goes through Google Vault's export process, real-time or near-real-time collection is not supported. Content captured during a sync reflects what was available in Vault at the time the export was created.

Sync Interruptions

If a sync is interrupted while an export is being processed by Google, Onna's workflow infrastructure preserves the export state and resumes from where it left off on the next sync cycle. No data is lost and no duplicate exports are created.

Content Format

Conversation content is ingested as structured text. Formatting is preserved to the extent provided by the Vault export.