Client meeting notes - 4/9

Both Ackley and Nikan (our clients) agree: we need to focus more on the user experience (UX).

To this end, Sonny's action item for the week is to start implementing (and hopefully get close to finishing) a user account interface for our site.  There are two main features this will give to our users.  The first is to allow users to save their position in a graph traversal and come back later using their login; this feature is crucial, because some QuestionStates may require the user to go away for some time as they fiddle with the car.  It's very important that they can come back and pick up right where they left off.  The second is to save the successful traversals (diagnoses) that a user has performed.  This feature is not as critical, but would be very nice for the UX.  This will give them a kind of service history for their car, showing issues that have come up with it and how they were resolved.  These could form the basis of a simple set of maintenance records for a user.

Alan is going to create a sort of demonstration that shows his graph algorithm in action.  This will help the UX by showing them the types of things MechanApp does in the background to help them that they don't necessarily see in the UI.  I think the end result of this task should be some kind of short video explaining the intelligence behind the system that can be watched from the MechanApp.net homepage.

David is working on backtracking still (along with Sonny), but it should be finished soon.  This is another critical function that will allow the user to cancel an answer they have selected and go back to the question again.  This is not only important if the user accidentally selects the wrong answer to a question, but it also will make the user feel MUCH more in control of the process, rather than some kind of semi-passive actor as they are now.

My main action item for the week is also related to the UX.  I am adding at least one more full graph to the database, maybe two.  Content is becoming an issue.  Our demos and pitch are going to quickly lose steam if we are always going over the same diagnostic, "won't start".  Adding some diagnostic procedures from different systems on a car (like brakes, cooling system, etc.) will show how flexible our system is.  It will help to show that our system is a general framework for auto diagnostics, and not just a 'bag of tricks' of some kind.  And it will definitely make the system more usable.  Alan even said his Jeep is having some cooling system issues, and I'm hoping that adding a chart related to that could allow us to make him and his car the guinea pig for a real-life usage :)