H Hypernovi customer docs Documentation space
Browse spaces
Customer documentationGuides

Creating Virtual Colleagues

Set up and configure Virtual Colleagues for your team.

Maintained by Hypernovi · Updated for the current product release

Creating Virtual Colleagues

This guide walks you through the full process of creating, configuring, and managing Virtual Colleagues in IntuitivePM. By the end, you will have an AI-powered team member running on a schedule and delivering reports to your preferred channels.

Prerequisites

Before creating a Virtual Colleague, make sure you have:

  • At least one project created in your company workspace.
  • Sufficient AI credits in your account (each colleague run consumes credits).
  • A role with write permission on Virtual Colleagues in the project you want to assign the colleague to (the built-in Admin and Super Admin roles have this by default).
  • (Optional) A Slack incoming webhook URL, Teams webhook URL, or company SMTP configured for notifications.

Step 1: Open the Creation Wizard

Navigate to Virtual Colleagues in the left sidebar. Click the Create Colleague button in the top right corner. This opens the creation wizard, which walks you through each configuration step.

Step 2: Basic Information

Name

Give your colleague a descriptive name that reflects its purpose. Good names help your team understand what the colleague does at a glance.

  • “Morning Standup Bot” — for a daily status reporter
  • “QA Gate Checker” — for a colleague that monitors task quality
  • “Sprint Analyst” — for weekly sprint performance reports

Avatar

Select an avatar for your colleague. This appears in activity logs, notification messages, and the Virtual Colleagues list. Choose something that visually distinguishes it from human team members.

Description

Write a brief description of what this colleague is responsible for. This is displayed on the colleague’s card and helps team members understand its purpose. For example: “Scans all active sprints every morning and posts a summary of overdue tasks to the #engineering Slack channel.”

Step 3: Role Selection

Choose one of the nine available roles. The role determines the colleague’s default chasing, reminder, escalation, reporting, and tone settings, and the type of analysis it performs.

RoleBest For
Scrum MasterSprint health, blocker detection, standups, stale-task sweeps
Project LeadMilestones, resource tracking, delivery reporting
Product OwnerBacklog health, prioritisation, release readiness
ManagerPeople, workload balance, capacity and performance
MemberAn individual’s own tasks, deadlines, and reviews
MyselfA private personal digest, focus and self-tracking
CustomerExternal milestone and delivery updates
Agent OrchestratorPlanning, delegating, and reviewing agent work
AgentExecuting assigned task or subtask work

Each role ships with pre-configured defaults optimized for that type of analysis. You can adjust the colleague’s behavior after creation by editing its configuration.

Role flexibility

A role sets defaults, it does not restrict what data the colleague can access. A “Member” role colleague can still report on sprint health — the role simply tunes how insistently it chases, who it escalates to, and how its output reads.

Roles differ in how much they chase

The roles are not interchangeable in tone. Scrum Master chases daily and escalates after two days overdue; Manager waits five days; Myself and Customer never chase anyone or escalate at all. Pick the role whose default insistence matches what you want, then fine-tune it.

Step 4: Schedule Configuration

Virtual Colleagues run on a schedule. Configure when and how often your colleague should execute.

Frequency

Choose how often the colleague runs:

Credit usage

Each scheduled run consumes AI credits. Hourly colleagues use significantly more credits than daily or weekly ones. Start with a lower frequency and increase it only after verifying the output is useful.

  • Hourly — Runs once every hour. Best for time-sensitive monitoring like overdue task alerts or QA gate checks.
  • Daily — Runs once per day at a specified time. Ideal for morning standups, daily digests, and end-of-day summaries.
  • Weekly — Runs once per week on a specified day and time. Good for sprint reports, weekly status updates, and capacity reviews.
  • Monthly — Runs once per month on a specified date and time. Suited for monthly retrospective data, utilization reports, and trend analysis.

Time of Day

For daily, weekly, and monthly schedules, set the time of day when the colleague should run. Times are displayed in your configured timezone.

Common patterns:

  • 8:00 AM for morning standup summaries (runs before your team’s standup meeting).
  • 5:00 PM for end-of-day status digests.
  • 9:00 AM Monday for weekly sprint reports (runs at the start of the work week).

Scheduling Examples

Here are concrete scheduling configurations and what they produce:

Every Monday at 9:00 AM   → Send sprint kickoff message to Slack
                             (lists tasks for the week, highlights critical items)

Every Friday at 4:00 PM   → Generate weekly progress report
                             (completed tasks, pending items, velocity summary)

Every day at 8:00 AM      → Morning standup digest to Teams channel
                             (what changed overnight, what is blocked, due today)

Every hour                → QA gate check
                             (flag any tasks that moved to Done without QA lane)

1st of each month, 9 AM   → Monthly utilization report via Email
                             (team workload balance, over-assigned members)

Tip: Start with a daily or weekly schedule when setting up a new colleague. Once you verify the output is useful, you can increase frequency or add more channels.

Timezone

Select the timezone for your schedule. The colleague’s run time is calculated based on this timezone, which is especially important for distributed teams spanning multiple time zones.

Step 5: Task Configuration

AI Analysis

Use AI Analysis when you want the colleague to provide planning and project-health insight. This includes stale task detection, overdue work, blockers, risks, useful summaries, and other planning signals.

AI Analysis does not have to be enabled for task execution. It is an insight feature, not the execution switch.

Auto-Act on Assignment

Use Auto-Act on Assignment when the colleague should act on work assigned to it. This is separate from AI Analysis.

When company orchestration is enabled, Auto-Act can be paired with an approved execution mode:

  • Featured AI services: use company-approved built-in AI providers.
  • External execution: use an admin-approved external agent runtime.

External execution appears only when your company administrator has enabled it and approved at least one external agent/runtime.

Personal Orchestrator

Some workspaces use a personal orchestrator Virtual Colleague. This colleague belongs to a member and can be customized as that member’s planning assistant.

You can favorite an orchestrator Virtual Colleague and open chat from the favorite context menu. Use that chat to plan work, ask questions, or review past planning conversations.

For the full workflow, see Agent Orchestration. Company administrators can use Set Up Agent Orchestration to configure orchestrators, agents, external runtimes, and Work Files permissions.

Define the scope of what your colleague monitors and analyzes.

Project Assignment

Select one or more projects for the colleague to monitor. The colleague will only have access to data within the assigned projects. You can assign a colleague to multiple projects if you want a cross-project summary.

Priority Rules

Optionally configure priority filtering:

  • All priorities — The colleague analyzes all tasks regardless of priority level.
  • Urgent and High only — The colleague focuses on high-priority items, useful for alert-style colleagues that should only flag pressing issues.
  • Custom filter — Select specific priority levels to include in the analysis.

Task Scope

Define which tasks the colleague should consider:

  • All active tasks — Includes every non-archived task in the assigned projects.
  • Current sprint only — Limits analysis to tasks within the active sprint.
  • Overdue tasks only — Focuses on tasks past their due date.

Step 6: Notification Settings

Configure where the colleague delivers its output. You can enable one or more channels.

Slack

Enter your Slack incoming webhook URL. The colleague will post formatted messages to the channel associated with that webhook.

To get a webhook URL:

  1. Go to your Slack workspace’s Apps settings.
  2. Search for Incoming Webhooks and add it.
  3. Choose the channel where you want reports delivered.
  4. Copy the webhook URL and paste it into the Slack field.

Use the Test button to send a sample message and confirm the webhook is working.

Microsoft Teams

Enter your Teams webhook URL. The colleague sends Adaptive Card messages to the connected Teams channel.

To get a webhook URL:

  1. In your Teams channel, click the three-dot menu and select Connectors (or Workflows for newer versions).
  2. Add the Incoming Webhook connector.
  3. Name it and copy the generated URL.
  4. Paste it into the Teams field.

Use the Test button to verify delivery.

Email

Enable email notifications to send reports to specified email addresses. Email delivery uses either your company’s configured SMTP settings or IntuitivePM’s default mail service.

Enter one or more recipient email addresses. The colleague will send a formatted HTML email with its report output at each scheduled run.

Step 7: Review and Create

Review all your settings on the summary screen:

  • Name, avatar, and description
  • Selected role
  • Schedule (frequency, time, timezone)
  • Assigned projects and task filters
  • Notification channels

Click Create Colleague to save. The colleague will appear in your Virtual Colleagues list and begin running on its next scheduled time.

Monitoring Colleague Activity

Once a colleague is active, you can monitor its performance from the Virtual Colleagues page.

Activity Log

Each colleague has an activity log that records every execution. For each run, you can see:

  • Timestamp — When the colleague ran.
  • Status — Whether the run completed successfully or encountered an error.
  • Output — The generated report or alert content.
  • Delivery status — Whether notifications were successfully sent to configured channels.

Last Run

The colleague card displays the time of its most recent execution and whether it succeeded. A green indicator means the last run completed successfully. A red indicator means the run failed — check the activity log for details.

My Colleague Messages Widget

Your dashboard includes a My Colleague Messages widget that aggregates recent outputs from all your active colleagues. This gives you a single view of what your AI team members have been reporting without navigating to each colleague individually.

Colleague Insights Widget

The Colleague Insights widget on your dashboard provides aggregate statistics: total runs this week, success rate, most active colleagues, and credit usage breakdown.

Monitoring Your Colleague

Beyond the widgets and logs above, here is a practical checklist for keeping colleagues healthy:

Where to find activity logs: Navigate to Virtual Colleagues in the sidebar, click on the colleague, and open the Activity tab. Each entry shows the run timestamp, success or failure status, the full generated output, and delivery confirmation for each channel.

How to pause and resume a colleague: Toggle the Active switch on the colleague’s detail page to pause it immediately. The colleague retains all its configuration while paused. Toggle it back on to resume — the next run will occur at the next scheduled time based on the current time and the configured frequency.

What to do if a message fails to deliver:

  1. Open the activity log and find the failed run. The delivery status column will show which channel failed.
  2. Click the entry to expand it and read the error message (for example, invalid_token for Slack or 403 Forbidden for Teams).
  3. Navigate to the colleague’s Notification Settings and verify the webhook URL is still valid. Slack and Teams webhook URLs can expire if the connected app is removed or reinstalled.
  4. Paste the URL into the webhook field again and click Test to confirm delivery before saving.

Tip: Always use the Send Test Message button after entering a webhook URL — before enabling the schedule. This confirms the integration is working and shows you exactly what the colleague’s messages look like in your channel.

To modify an existing colleague:

  1. Navigate to Virtual Colleagues in the sidebar.
  2. Click on the colleague you want to edit.
  3. Update any fields — name, role, schedule, projects, notifications.
  4. Click Save Changes.

Changes take effect on the next scheduled run. If you change the schedule, the next run time recalculates based on the new configuration.

Activating and Deactivating

You can temporarily deactivate a colleague without deleting it. Deactivated colleagues retain their configuration but do not run on schedule. This is useful during quiet periods or when you want to pause a colleague while adjusting its settings.

Toggle the Active switch on the colleague’s detail page or card to activate or deactivate.

Deleting a Colleague

To permanently remove a colleague:

  1. Hover over the colleague card to reveal the Delete button.
  2. Click Delete and confirm the action.
  3. The colleague and its activity history are permanently removed.

Deletion is irreversible. If you might need the colleague again, consider deactivating it instead.

Tips for Effective Colleagues

  • Start with one colleague and monitor its output for a week before creating more. This helps you calibrate expectations and refine the configuration.
  • Match frequency to urgency. Hourly is for alerts; daily is for summaries. Running a detailed report hourly wastes credits and floods your channels.
  • Use descriptive names. Your team will see colleague names in Slack, Teams, and email. “Sprint Health — Engineering” is clearer than “Bot 1.”
  • Review outputs regularly. AI-generated content is helpful but not infallible. Treat colleague reports as starting points, not final judgments.
  • Monitor credit usage. Each run consumes AI credits. Check your credit balance in Settings and adjust schedules if usage is higher than expected.

Next Steps

On this page