Scale B.O.S.S. Settings Guide
AI Data Management

How to enable AI intentionally, curate training data, and govern AI chat retention.

Path: Settings > Integrations & AI > AI Data Management

Scope: Organization-level AI controls, training content, and chat retention

What this page is

AI Data Management is the organization-level control center for how Scale B.O.S.S. uses AI in your environment. This is where leadership decides whether AI is active, whether it may access organization data, which capabilities are enabled, what documents or links train the assistant, and how chat history is retained or purged.

This is a governance page, not a personal preference page. Choices made here affect AI behavior across the entire organization.

Why it matters A good AI setup balances capability, privacy, relevance, and control. The goal is not simply to turn AI on — it is to make AI useful, governed, and aligned with how the organization actually operates.

Workspace overview

AI Data Management covers three administrative responsibilities: feature governance, training inputs, and chat-history control.

At a glance: what each area controls

Section	What it controls	Why it matters
AI Features	Master enablement, data-access consent, and individual AI capability toggles	Controls whether AI is available and what it can perform
AI Training Data	Documents and links used to train the AI assistant	Improves relevance and organization-specific accuracy
AI Chat Data Management	Retention, chat statistics, and purge actions	Supports privacy, cleanup, and administrative oversight

1. AI Features

The AI Features card is the command center for organization-wide AI behavior. It contains the master AI switch, AI Data Access consent, and capability toggles that determine what AI can do inside the platform.

Treat this section as policy, not convenience. Settings here affect the entire organization.

Master AI switch

The upper-right toggle is the broadest control on the page. When enabled, the organization can use AI capabilities. When disabled, AI is turned off at the highest level.

AI Data Access

This row controls whether AI may access organization data to produce insights. The page records who granted consent and when, creating a visible audit trail that makes data approval explicit rather than implied.

Feature controls

AI B.O.S.S.: Strategic AI sidekick and coaching assistant.

AI Insights: KPI analytics and dashboard insight generation.

Messaging AI: Message refinement, smart replies, and summaries.

Weekly Tips: My Dashboard AI insights and nudges.

Financial Docs: AI document processing for financial material.

Notes AI: Note summarization and to-do extraction.

Best practice Do not enable every AI feature by default. Start with the capabilities the organization is ready to govern, then expand as usage and training quality mature.

2. AI Training Data

The AI Training Data card is where the organization provides the assistant with source material to learn from. This is what makes AI more useful in your specific environment rather than in a general one.

The training library is organized around Documents and Links, even before any content has been uploaded.

How uploads work

The upload area supports drag-and-drop and click-to-browse file selection. Supported formats include PDF, Word, TXT, CSV, and Excel, with a maximum file size of 10MB.

What belongs here

Upload current, organization-specific material that will genuinely improve AI responses: SOPs, playbooks, role guides, strategic frameworks, approved process documents, and reference material your team actively uses.

Avoid uploading outdated, duplicative, contradictory, or irrelevant files. Low-quality content reduces AI response clarity and accuracy.

What to review periodically

Whether the library reflects how the organization currently operates.
Whether the content is current enough to trust.
Whether document quality is high enough to improve — not undermine — AI answers.

3. AI Chat Data Management

The AI Chat Data Management card governs the retention and cleanup of AI conversation history. Administrators can view volume, review the retention policy, and trigger an immediate purge when needed.

The purge tool applies to chat history only — not to training data or account configuration.

Retention policy

AI chat conversations are automatically archived after 30 days and then permanently deleted. Training data, knowledge resources, and account settings are retained for as long as the organization is active. This distinction is critical: chat history and training content are managed separately.

Snapshot metrics

The Conversations, Messages, and Last Purge cards provide a quick operational snapshot, helping administrators gauge chat activity volume and confirm whether a recent purge has occurred.

Purge Now

The Purge Now action immediately deletes all AI chat conversations for the organization. Training data is not affected. Use this as a deliberate governance action, not a routine operation.

Use purge carefully This action is best suited for privacy cleanups, testing resets, or governance reviews. It is not a substitute for proper AI policy or training-data curation.

What good administration looks like

AI is enabled intentionally, not casually.
AI Data Access has been approved knowingly and is traceable to a visible consent record.
Only the AI capabilities that fit the organization's operating model are turned on.
Training data is curated like a knowledge base, not uploaded indiscriminately.
Administrators understand the difference between chat-history retention and training-data retention.
Purge authority is treated as a governance control.

Common mistakes to avoid

Enabling every AI feature before defining what the organization wants AI to do.
Expecting AI quality to improve without providing high-value training material.
Uploading outdated or contradictory documents into the training library.
Assuming chat purges remove training data or system knowledge.
Treating this page as a one-time setup rather than an ongoing governance responsibility.

Final guidance Use AI Data Management to make AI useful, governed, and relevant. Enable only what the organization is ready to support, provide the assistant with high-quality knowledge, and treat chat retention and purge settings as real administrative responsibilities.

Scale B.O.S.S. AI Data Management Settings Guide