Skip to content
Onplana Docs

Generate a project plan with AI

All plans Member

Instead of starting from a blank page, you can describe what you want to deliver and let Onplana draft the plan: a structured set of tasks with estimates, grouped sensibly, plus starter risks worth tracking. You then edit the draft like any other project.

Where you can generate a plan

Section titled “Where you can generate a plan”
  • During onboarding: the AI kickstart step when you create your first organization offers to draft your first project from a description.
  • From the AI Project Assistant on the dashboard: use the Intake action and describe the project.

Generate and review

Section titled “Generate and review”
  1. Describe the project in two or three sentences: what you are delivering, roughly for whom, and any known deadline.
  2. Onplana drafts the plan: tasks with estimated hours, subtasks where the work splits naturally, and a few starter risks.
  3. Review the draft. Everything is editable, and nothing exists in your workspace until you accept it.
  4. Accept to create the project. The project, its tasks, and the risks are created together in one step.

Add a plan to a project you already have

Section titled “Add a plan to a project you already have”

A generated plan doesn’t have to become a new project. Ask the assistant (or a connected agent) to add the plan to an existing project, and the tasks, epics, milestones, and risks land inside the project you name instead of a fresh one.

Adding is strictly additive: it never renames the project, never changes its owner, and never moves the dates of anything already there. New tasks are appended after your existing ones, so the plan slots in without disturbing work in flight.

FAQs

Section titled “FAQs”

Can I regenerate if I do not like the draft? Yes, adjust your description and generate again. Drafts cost AI tokens only when generated, so iterate on the wording rather than accepting a weak plan.

Does the AI plan replace templates? They complement each other: templates are exact and repeatable, AI plans are tailored to your description. For work you run the same way every time, a template is usually better.

Best practices

Section titled “Best practices”
  1. Include concrete constraints in the prompt. Launch date, team size, tech stack, budget. Specifics yield specifics.
  2. Iterate on the description, not the draft. Cheaper to regenerate from a refined prompt than to manually fix 30 misnamed tasks.
  3. Treat the AI plan as a starting point. Edit names, dates, and estimates after acceptance. The draft is rarely perfect.
  4. Combine with templates for hybrid workflows. AI-generated structure
    • template-driven repeatable phases = the best of both.
  5. Cite the AI-generated origin in the project description. Helps future readers understand why some tasks may look templated.

Troubleshooting / common pitfalls

Section titled “Troubleshooting / common pitfalls”
  • Draft is too generic. Description too vague. Add concrete constraints.
  • Generated plan exceeds budget. AI doesn’t know your budget unless you mention it. Include in prompt.
  • Tasks have no dependencies. AI plan focuses on structure, not sequencing. Add dependencies after acceptance.
  • AI rejected the prompt. Prompt too short or vague. Provide context.
  • Plan token cost too high. Cap description length; explore via chat first.

How this combines with other features

Section titled “How this combines with other features”

Coming from another tool

Section titled “Coming from another tool”
ToolMapping
Microsoft Project AI featuresDirect concept
Asana AI smart fieldsAdjacent
ChatGPT for planningOnplana adds first-class structure
Custom OpenAI promptOnplana has it built-in
Smartsheet AIDirect
Section titled “Related”