All articles

AI Project Risk Prediction

Table 45. Availability - Legacy plans


AI Project Risk Prediction monitors your projects for any risks that may cause you to miss your deadlines, then assigns a risk level (low, medium, or high). AI estimations have replaced the previous system of Project Health calculations (color-coding). You can find these estimations in the Project Progress window in the project info panel, Table view, the project dashboard widget, and project-based reports.

Project risk assessment takes into account a number of factors including:

  • project complexity (the number of tasks, subtasks, subfolders, and subprojects, and how they are structured)

  • the number of completed and overdue tasks

  • the number of assignees

  • how many active and overdue projects the owner has

  • task activity

  • the owner’s previous projects (if they were not completed on time this may increase the predicted risk)


Project risk information can also be shared with a user via Wrike's Automation engine.

Important Information

To get risk estimates, you'll need to Turn on Project Progress on necessary projects.

Previously, project health was based on a simple calculation, projects created before 18th November 2021 will continue to use project health calculations until the end of the project. Using AI Project Risk Prediction means a machine learning approach gives you a more comprehensive risk estimate based on dozens of current and historic factors associated with your projects.

Keep in mind that the assessment of project progress (a percentage) remains the same. The color-coding is project-based. All users will see either the previous project health status or a new version with risk estimation if it's enabled for their projects.

If the AI finds risks, it will give you 1 – 3 summaries of them in the Project Progress window, so that you can look into these potential issues and address them (for example, it might tell you 12 tasks overdue.) Projects which have just been started might show the explanation Preliminary risk assessment, because not enough data is available yet to make a risk prediction. Conversely, projects which have been running for over 500 days will show the explanation Project is too long for AI to calculate project risk, and risk will not be calculated.


Understand your Project Risk

You'll only get project risk estimates for:

  • Active projects which include criteria which allow for risk calculation such as duration, start date, end date, etc. (subprojects and tasks do not need to have start and end dates).

  • Projects with Project Progress activated

The more accurate and up to date the details of your projects are (start and end dates, status, project's tasks info, etc), the more accurate the risk predictions will be. The project risk is calculated each time you make changes to your projects or tasks included in your projects, and is visible after a couple of seconds.

Projects which don't meet these criteria are labelled Not assessed and marked in grey.

Your Project Risk Prediction consists of a level (low, medium, or high) and a corresponding colour which can be seen in your Project Health window:

  • Low risk - Green

  • Medium risk - Amber

  • High risk - Red

What's next?