Virtual Colleagues Overview

Virtual Colleagues are AI-powered team members that live inside your IntuitivePM projects and automate recurring work. They operate on schedules you define, performing tasks like scanning boards for overdue items, generating status reports, running QA checklists, and sending summaries to Slack, Teams, or Email. Think of them as always-on teammates that handle the repetitive parts of project management so your human team can focus on building.

What Makes Virtual Colleagues Different

Unlike one-off AI prompts, Virtual Colleagues are persistent. Once created, a colleague has a name, a role, an avatar, and a defined set of responsibilities. They run on a schedule — hourly, daily, weekly, or monthly — and produce outputs that are logged, reviewable, and delivered to the channels your team already uses.

Each Virtual Colleague is scoped to specific projects and tasks. They read your board state, analyze task metadata (status, priority, assignees, due dates, estimated and logged hours), and take action based on their role. Their activity is logged so you always have a record of what they did and when.

Virtual Colleagues can also participate in agent orchestration when your company has enabled it. In that mode, a personal orchestrator Virtual Colleague can help plan work, assign approved execution agents, review results, and answer planning questions from chat or task comments.

Available Roles

IntuitivePM offers nine built-in roles. A role is a starting preset, not a fixed personality: choosing one sets the colleague’s default chasing, reminder, escalation, reporting, and tone behaviour, and you can override any of it afterwards.

Roles differ mainly in who the colleague works on behalf of and how insistent it is — a Scrum Master chases every day and escalates after two days overdue, while a Customer never chases at all.

Role Comparison

Scrum Master

The most active role. Chases open work daily, reminds ahead of and after due dates, escalates to the Product Owner after two days overdue, and flags tasks sitting in the same status for more than two days. Produces a daily report covering velocity, completion rate, and overdue, blocked, and stale counts.

Project Lead

A calmer cadence aimed at delivery tracking. Chases every other day, escalates to a Manager after three days overdue, and treats a task as stale after five days in the same status.

Product Owner

Watches the backlog rather than the sprint board — priorities, readiness, and release scope — escalating to a Manager when items slip.

Manager

The least frequent chaser of the management roles. Focuses on people and workload, escalating to senior management after five days overdue. Useful for spotting overloaded team members before burnout.

Member

Scoped to one person’s own work: their tasks, deadlines, and reviews. Notifies in-app by default and escalates to the Project Lead when something is stuck.

Myself

A personal, self-tracking digest. It never chases anyone and never escalates — it just reminds you about your own upcoming and overdue work.

Customer

Built for external stakeholders. Chasing, reminders, and escalation are all off; it delivers milestone and delivery updates by email only.

Agent Orchestrator

Plans work, delegates it to approved execution agents, reviews the results, and coordinates the overall flow. This is the role used to drive Agent Orchestration.

Agent

Executes the task or subtask work assigned to it, within the permissions its company has approved.

Use Cases

Virtual Colleagues are most effective for work that is repetitive, time-sensitive, and data-driven. Here are common patterns:

• Automated daily standups — A Scrum Master colleague scans all active tasks each morning and posts a summary to Slack with what changed overnight, what is blocked, and what is due today.
• Weekly delivery reports — A Project Lead colleague generates a milestone and velocity summary every Monday morning, delivered to your Teams channel.
• Stale-work sweeps — A Scrum Master colleague runs hourly, flagging tasks that have sat in the same status for too long or moved to Done while still blocked.
• Backlog grooming — A Product Owner colleague reports on unprioritised or stale backlog items before each planning session.
• Personal focus digest — A Myself colleague emails you a private summary of your own upcoming and overdue work, without notifying anyone else.
• Workload balancing — A Manager colleague generates weekly team utilization reports so managers can spot overloaded team members before burnout sets in.
• Customer-facing updates — A Customer colleague emails an external stakeholder a milestone summary without ever chasing your team.

How They Interact with Your Data

Virtual Colleagues read from your project’s tasks, epics, sprints, and member data. When a colleague runs, it:

1. Queries the current state of assigned projects and tasks.
2. Applies role-specific analysis (overdue detection, sprint progress calculation, QA coverage checks).
3. Generates a structured output using AI.
4. Delivers the output to configured channels (Slack, Teams, Email).
5. Logs the activity with timestamp, status, and output details.

By default, Virtual Colleagues focus on analysis, reports, reminders, and communication. When Auto-Act on Assignment or agent orchestration is enabled, approved colleagues can also work on assigned tasks or subtasks within the permissions configured by your company.

AI Analysis and Auto-Act are separate settings:

• AI Analysis helps with project insights such as stale work, overdue tasks, blockers, risks, and useful planning summaries.
• Auto-Act on Assignment allows a colleague to execute assigned work when the company and member permissions allow it.

See Agent Orchestration for the full task, review, and Work Files workflow. If you are configuring the workflow for a company, start with Set Up Agent Orchestration.

Limitations

Virtual Colleagues are powerful but have clear boundaries:

• They cannot approve pull requests or merge code. They analyze project management data, not source control.
• They work within set schedules. A colleague configured for daily runs will not react to events in real time. Use webhooks if you need instant notifications.
• Outputs need human review. AI-generated summaries and reports should be reviewed before acting on them, especially for customer-facing communications.
• They require project access. A colleague can only analyze projects it has been assigned to. It cannot see data from other projects or companies.
• AI credits are consumed per run. Each colleague execution uses AI credits from your account balance. Monitor your credit usage in Settings.
• External execution requires admin approval. A colleague can use an external runtime only when a company administrator has approved the agent identity and runtime.
• They act within configured permissions. A colleague can work only with the projects, tasks, files, and channels allowed by its configuration.
• Review remains required for agent-executed work. An orchestrator or human reviewer accepts work to Done or sends it back with feedback.
• Each colleague requires at least one configured communication channel. To receive output, a Slack webhook, Teams webhook, or email address must be set up. A colleague with no notification channels will run but its output will only appear in the activity log.

Getting Started

Ready to create your first Virtual Colleague? Head to the Creating Virtual Colleagues guide for step-by-step setup instructions, including role selection, schedule configuration, and notification channel setup.

For agent-based work, continue with:

• Set Up Agent Orchestration
• Assign Work To Agents
• Connect External Agent Runtimes