The AI Project Builder

The AI Project Builder is the AI-assisted planning tab inside each Farline AI project plan. It lets you describe your project in plain language and generates a structured plan — complete with workstreams, work items, team size, and a timeline estimate — ready to forecast.

What It Does

Instead of manually entering every work item and workstream, you describe your project the way you would explain it to a colleague:

"Plan a mobile app redesign with a team of 3 engineers and 1 designer over 10 weeks. We need design, development, and QA phases."

The AI produces a plan that includes:

• Named workstreams (e.g., Design, Frontend, QA)
• Work items within each workstream with relative size estimates
• A suggested team allocation and start/end timeline
• A Gantt chart forecast based on those assumptions

How to Use It

Opening the AI Builder

1. Open the Farline AI tab in a Jira project
2. Create a project plan, or open an existing one
3. Select the AI Builder tab
4. You'll see a chat interface with starter prompts — pick one or type your own description

Starter Prompts

The interface offers starter prompts to get you going:

• "Plan a product launch"
• "Plan a migration or infrastructure project"
• "Plan a design and build project"

These are starting points — you can modify any of them or ignore them and describe your project from scratch.

Iterating on the Plan

After the AI generates an initial plan, you can refine it:

• Add or remove workstreams: "Remove the QA workstream" or "Add a marketing phase"
• Adjust team size: "Change to 5 engineers"
• Adjust timeline: "Make this a 6-week plan instead"
• Change scope: "Add work items for performance testing"

The AI updates the plan and re-renders the forecast with each change.

Applying Changes

Each successful AI Builder turn updates your project definition. Once the definition is valid, the forecast refreshes in the Forecast Results and Charts tabs.

What the AI Produces

Internally, the AI generates a YAML-structured project plan that specifies your workstreams, work items, and team configuration. This plan is rendered as a forecast scenario — you don't need to read or edit the YAML directly. If you want to make fine-grained changes later, you can edit work items individually in the project editor.

Connecting to Jira

The AI Builder can work alongside Jira data in two ways:

• It can help you shape a plan after you import existing Jira epics into Farline AI
• After you've built a plan, Sync to Jira can update mapped Jira fields and create missing epics for manual work items

This requires the write:jira-work permission — in plain terms, the app needs permission to update Jira epics on your behalf.

Tips

• Be specific about team size and duration. The AI uses these as the core inputs for timeline estimation. Vague inputs produce vague plans.
• Iterate freely. The AI remembers the conversation context, so you can say "make the design phase shorter" without repeating the whole project description.
• Use /help if you're unsure what the AI Builder supports. The built-in slash commands include scenario-copying and other editing shortcuts.
• You can always edit manually. The AI Builder produces a starting point. Add, remove, or adjust work items in the editor at any time.

Limitations

• The AI Builder works best for projects with 2–8 workstreams and up to ~40 work items. Very large projects may produce estimates that need manual refinement.
• The AI generates timeline estimates based on team size and relative work item sizes — it does not account for individual engineer velocity or Jira story points.
• Historical Jira data (past sprints, velocity) is not used as input in the current beta release.

Related

• Scenario Forecasting — create and compare forecast scenarios after building your project
• Syncing with Jira — push forecast dates back to your Jira epics