The Risk Solver App is used to create and run Monte Carlo simulation models, and view results -- charts and statistics -- from the simulation process.

Important to know:
 
-- The Risk Solver App will work in both Excel 2013 and the Excel Web App / Excel Online, but its backend server is designed to run simulations for models (only) in Excel workbooks that are stored in a SharePoint document library.  (If an Excel workbook is stored on a user's local hard disk, the backend server has no way to access it.  The Risk Solver App detects this, and advises the user of the issue.)
 
-- Because Excel Online currently lacks an "Insert App" command, to start you must use Excel 2013 to insert the Risk Solver App into a workbook, then save the workbook to a SharePoint document library.  When you next open the workbook in the Excel Web App, the Risk Solver App will appear, at the place where it was embedded.

-- Because Excel Online does not save workbook changes automatically, after defining Distributions or Outputs with the Risk Solver App (or making other changes in the model), you must use File Save As to save your workbook before you click the green arrow to run a simulation.  Tooltip help on the green arrow reminds you to do this.

TESTING THE RISK SOLVER APP

You can perform a simple test by inserting the Risk Solver App in a blank workbook, saving this workbook to a SharePoint document library, and starting the app, which opens a Task Pane.  If you click the button Insert Example at $A$1, the App will create a complete Monte Carlo simulation example model on the active worksheet.  This is an Airline Revenue Management model, illustrated and discussed in the video, with one probability distribution (for passenger "no shows") and one output (for total revenue).  Just save the workbook and click the green arrow to run the simulation.

For a second test model, the College Fund model also illustrated and discussed in the video, click this link to download the Excel workbook, then upload this workbook to a SharePoint document library.  You can also create your own test models.