Sending an AI prompt in Maverick isn't just a chat exchange — it triggers a chain of scheduling, visualization, and data updates across the project. Here's what happens from the moment you confirm an AI update to when the screen settles.
1. Task details change in the task grid immediately
When you confirm an AI update, changes appear in the task grid right away. Task names, dates, durations, status values, and assigned resources all update in real time — no page refresh needed. The change history log records exactly what the AI modified, who triggered the update, and when it happened.
2. The Gantt chart redraws to reflect the new schedule
After an update, switch to the Gantt view to see the new timeline visually. Task bars move to their new positions, dependency arrows redraw, and the critical path recalculates automatically. The Gantt is the fastest way to verify that date changes make sense in context — a shifted task that looks fine in a date column might reveal a gap or overlap on the chart.
3. Dependent tasks cascade to new dates automatically
If the AI moves a predecessor task, all downstream tasks with Finish-to-Start links shift forward or backward automatically. The scheduling engine handles the cascade — you don't need to manually update every task in the chain. This is one of the biggest time-savers of AI-driven scheduling: a single prompt can correctly reschedule a dozen linked tasks in one operation.
4. The resource allocation bar chart updates in parallel
Any resource reassignments or workload changes the AI makes are reflected immediately in the allocation bar chart. Over-allocated resources turn red, under-utilized ones turn amber, and balanced resources show green. Opening the allocation chart after an AI update is the quickest way to catch staffing conflicts before they become real problems.
5. You can ask the AI to revise its own changes immediately
If the result isn't quite right, type a follow-up prompt: "Move the testing phase to the last week of the month instead." The AI reads the current state of the project — including its own prior changes — and refines the schedule based on your correction. You can iterate as many times as needed in the same session without losing context.
6. Nothing is committed until you explicitly approve the update
Maverick shows you a preview of AI changes before applying them. You review the proposed modifications, approve what you want, and discard anything that looks wrong. This review step is your safety net against AI hallucinations reaching your live project data — you are always in control of what actually gets saved.