Team Chat Environments & Collaboration

Keeptrusts supports team-scoped chat environments that provide isolated workspaces for different groups within your organization. Each team can have its own policy configurations, knowledge assets, conversation history, and access controls.

Use this page when

• You are setting up team-scoped chat workspaces with isolated policies and knowledge assets.
• You need to configure shared conversation visibility and role-based access controls for chat.
• You want to assign different gateway configurations and system prompts per team.
• You are managing team membership, roles, and cost allocation for chat usage.

Primary audience

• Primary: Platform Administrators creating team environments, Team Leads managing chat access
• Secondary: Compliance Officers verifying team isolation, Technical Leaders planning multi-team rollout

Team Chat Architecture

Team-scoped chat builds on the Keeptrusts identity model:

• Organizations contain one or more teams.
• Each team can be assigned specific gateway configurations and knowledge assets.
• Chat sessions inherit the team context of the authenticated user.
• Gateway keys (kt_gk_...) are scoped to the user's team membership.

When a user sends a message in the Chat Workbench, the gateway resolves the effective policy chain based on the user's team, ensuring each team operates within its own governance boundaries.

Setting Up Team Chat

Creating a Team

1. Navigate to Settings → Teams in the management console.
2. Click Create Team.
3. Provide a team name, description, and optional metadata.
4. Save the team.

Assigning Members

1. Open the team detail page.
2. Click Add Members.
3. Search for users by name or email.
4. Select a role for each member (see Role-Based Chat Access).
5. Confirm the assignments.

Assigning Gateway Configurations

Each team can be associated with a specific gateway configuration:

1. Navigate to Configurations in the console.
2. Select or create a configuration for the team.
3. In the configuration settings, bind it to the team.
4. Deploy the configuration to the relevant gateway.

Team members chatting through this gateway will have the team's policies applied automatically.

Shared Conversation History

Team environments support shared conversation visibility:

How Sharing Works

• By default, conversations are restricted to the owner and any explicitly added collaborators.
• The chat workbench header exposes a Share button in the compact chat header.
• If you click Share before sending the first message, Keeptrusts reserves a shareable conversation session so you can grant access or copy a link before the thread has content.
• Owners can add individual users or teams, choose collaborator or observer access, and copy a deep link to the current conversation.
• The General access control in the share dialog can switch a conversation between Restricted and Organization access.
• Individual messages within shared conversations remain attributable to the originating user.

Sharing a Conversation

1. Open a conversation in the Chat Workbench.
2. Click Share in the chat header.
3. Add a user or team and choose Collaborator or Observer.
4. If you want org-wide visibility, change General access from Restricted to Organization.
5. Use Copy link to share the direct conversation URL.

Viewing Shared Conversations

In the Chat Workbench:

1. Open the conversation sidebar.
2. Browse or search for conversations that were shared directly with you, your team, or your organization.
3. Click a conversation to view the full thread.

Shared conversations are read-only for observers. Collaborators can continue the thread in the same conversation.

Dedicated Tasks Workspace

The console sidebar now exposes Tasks as a dedicated destination alongside Chats and History.

• Signed-out visits to /tasks redirect to console sign-in and return to the same workspace after authentication.
• The /tasks page uses the same compact embedded shell as chat, but the drawer lists reusable tasks instead of conversation history.
• If reusable tasks already exist, opening /tasks lands on the most recently updated task thread by default so teams can resume recent automation work immediately.
• If no reusable tasks exist yet, /tasks opens the embedded task builder directly so teams can create their first automation without starting from a chat transcript first.
• The inline task builder keeps the form compact: enter a task name, describe the outcome, choose an agent, then use the searchable Where should it run? dropdown to pick one of that agent's ready run locations.
• Selecting a task opens its task thread with run events, approvals, notes, and connector-backed execution details in one place.
• Legacy detail URLs such as /tasks/{id} and /tasks/{id}/runs/{rid} resolve back into the same embedded /tasks workspace for the selected task.
• The task thread header uses the same Share button and popup pattern as chat, replacing the older inline task-collaborator rail.

Live Collaboration Signals

• When another collaborator is typing in the same conversation, the composer shows a compact avatar strip above the input.
• The typing indicator is best-effort and clears automatically when the other user stops typing or becomes idle.

Role-Based Chat Access

Keeptrusts provides granular role-based access for chat functionality:

Role	Capabilities
Chat User	Send messages, view own conversations, access bound knowledge assets
Chat Contributor	All Chat User permissions plus share conversations, contribute to team threads
Chat Moderator	All Contributor permissions plus review escalations, manage shared conversations
Chat Admin	All Moderator permissions plus configure team chat settings, manage knowledge bindings

Assigning Chat Roles

Roles are assigned at the team level:

1. Navigate to Settings → Teams → [Team Name] → Members.
2. Click the role dropdown next to a member.
3. Select the appropriate chat role.
4. Save changes.

Role Inheritance

• Organization-level roles provide baseline permissions.
• Team-level chat roles add or restrict capabilities within the team context.
• A user who is a Chat Admin in Team A may be a Chat User in Team B.

Team Knowledge Bases

Each team can maintain its own set of knowledge assets:

Creating Team Knowledge Assets

1. Navigate to Knowledge Base in the console.
2. Click Create Asset.
3. Select the owning team from the team dropdown.
4. Upload or author the knowledge content.
5. Promote the asset through the review lifecycle.

Knowledge Asset Visibility

Scope	Visibility
Organization-wide	Available to all teams
Team-scoped	Available only to the owning team
User-scoped	Available only to the creating user

Team-scoped assets are bound to gateways that serve that team. Organization-wide assets can be bound to any gateway.

Cross-Team Knowledge Sharing

To share a team asset with another team:

1. Open the asset detail page.
2. Click Manage Access.
3. Add additional teams to the access list.
4. Save the updated access settings.

The asset remains owned by the original team but becomes available for binding by the added teams.

Collaboration Workflows

Team Review Workflow

1. A Chat User creates a conversation exploring a topic.
2. The user shares the conversation with the team.
3. A Chat Moderator reviews the conversation for accuracy.
4. The moderator adds annotations or flags specific responses.
5. The team uses the reviewed conversation as a reference.

Escalation Within Teams

When a policy escalation occurs in a team conversation:

1. The escalation is routed to Chat Moderators within the team.
2. Moderators receive a notification in the console.
3. The moderator reviews the flagged content and approves or denies it.
4. The decision is logged in the audit trail with the moderator's identity.

Monitoring Team Chat Activity

Team administrators can monitor chat activity through the console:

• Usage dashboards: Token consumption, message counts, and active users per team.
• Policy trigger rates: How often team members encounter policy blocks or escalations.
• Knowledge citation rates: How frequently team knowledge assets are referenced.
• Cost allocation: Chat costs attributed to the team's wallet.

See Chat Analytics & Usage Insights for detailed analytics guidance.

Best Practices

Practice	Why It Matters
Use team-scoped configurations	Isolates policy boundaries between groups
Assign the minimum required chat role	Follows least-privilege access principles
Enable opt-in sharing by default	Balances collaboration with privacy
Maintain team knowledge assets actively	Keeps grounded responses accurate and current
Review escalation patterns regularly	Identifies training needs or policy gaps
Monitor per-team cost allocation	Prevents unexpected spend from any single team

Next steps

• Track team-level chat metrics in Chat Analytics & Usage Insights.
• Export team conversations for compliance in Chat Export for Compliance & Audit.
• Compare model performance across teams in Multi-Model Chat Comparison.

For AI systems

• Canonical terms: team chat, team-scoped environment, shared conversations, role-based access, gateway configuration binding, team knowledge base, visibility scope, cost allocation.
• Console pages: Settings → Teams (create/manage teams), Configurations (bind to teams). Gateway keys scoped to team membership.
• Roles: viewer (read shared conversations), member (send messages), admin (manage team settings, enable sharing).
• Best next pages: Chat Analytics, Chat Export, Knowledge-Grounded Chat.

For engineers

• Create teams in Settings → Teams, then assign members with roles (viewer, member, admin).
• Bind a gateway configuration to each team so the team’s policies are applied automatically when members chat.
• Shared conversations are opt-in at the conversation level — owners use the Share dialog to invite users, teams, or their full organization.
• Gateway keys inherit team scope from the authenticated user; no manual team binding is required on the key.
• Monitor per-team metrics in the Dashboard — token consumption, message counts, policy triggers, and cost.

For leaders

• Team environments provide governance isolation — each team operates within its own policy boundaries and cost allocation.
• Shared conversation visibility enables collaboration while maintaining individual message attribution.
• Role-based access follows least-privilege principles — not every team member needs admin or sharing capabilities.
• Per-team cost allocation supports showback/chargeback models for AI spending.
• Team environments scale organizational rollout — start with one team, validate policies, then expand to others.