Skip to main content
The AI for Work Assistant is your workspace hub for executing tasks, managing workflows, and coordinating across connected apps using natural language.

​
Getting Started

Log in to the Platform. The AI for Work Assistant is available on the homepage. Type an open-ended question or instruction in the Ask or search anything search box. For example, if you’re connected to Jira, type Get my Jira issues — the Jira agent retrieves your issues automatically.

​
Agentic Flow

The Platform manages complex, multi-step workflows by organizing tasks into a structured sequence, each handled by a specialized agent. Each agent performs a unique function, executes in the correct order, and passes information seamlessly to the next — enabling multiple instructions to complete with minimal manual effort. Key capabilities:
  • Coordinate multiple agents across connected apps in a single instruction.
  • Execute tasks sequentially with automatic handoff between agents.
  • Edit or delete individual actions before starting the flow.
  • Review and approve outputs (such as draft emails) before final execution.
Use when: You need to chain tasks across multiple apps or automate a multi-step workflow in one request.

​
Agentic Flow Example

Instruction: “Get my Jira issues. Let me know John Doe’s availability for a meeting tomorrow. Send the Jira issues to John Doe.” Agents involved:
  1. Jira — Fetches the user’s Jira issues.
  2. Google Calendar — Checks John Doe’s availability for a specified time.
  3. Gmail — Sends the fetched Jira issues to John Doe via email.
agentic-flow Steps to execute the flow:
  1. Enter your instructions and press Enter. Each action is mapped to its respective agent.
  2. Hover over any instruction to display the Edit and Delete icons to modify or remove individual actions.
  3. Click Start to initiate the flow. Tasks execute sequentially.
  4. The first two actions run automatically.
  5. For the third action, a draft email is created. Review and edit the draft, then click Send to complete the task.

​
File Storage and Data Handling

When you upload files during chat sessions, the platform can store files in managed cloud storage (AWS S3) under a multi-tenant architecture. This option is enabled by default but can be turned off by an administrator. This applies to:
  • All file uploads in Platform chat sessions.
  • Documents processed through Search AI functionality.
Disable or Enable S3 Storage Administrators can control whether uploaded files are persisted to S3 cloud storage.
  1. Go to Admin Hub → Admin Settings → Data Settings.
  2. Locate the Cloud File Storage (S3) toggle.
  3. Set the toggle to On (default) to store uploaded files in managed S3 storage, or Off to disable cloud persistence.
When S3 storage is disabled, uploaded files are available only for the duration of the active chat session and are not retained after the session ends.
Files uploaded during chat sessions are stored in the platform’s cloud infrastructure.
Data security: All files in the platform’s managed storage are protected with industry-standard encryption to ensure data confidentiality and integrity.

​
Uploading Files

Attach files to your conversation by clicking the attachment icon in the compose bar or by dragging and dropping a file directly into the chat interface. To upload using drag and drop:
  1. Drag a file from your local system over the chat window.
  2. The drop zone highlights to confirm the file can be released.
  3. Drop the file. The upload initiates automatically.
The upload progress indicator behaves consistently for both drag-and-drop and manual uploads. For information on supported file types and storage, see File Storage and Data Handling.