In the previous post, I introduced a scenario for discussing personal assistants.  In this post, I explore how personal assistants are useful at a tool for connecting the primary care physician to the back care specialist.

(Update: this post has had the terms changed to align with: Not an Agent, but a Personal Assistant.)

The Primary Care Process

What does a primary care physician, Betty, do?  Many things, and for this scenario we focus on how they diagnose a problem that the patient has.  There are many problems that they detect and address immediately.  There are also regular tests (e.g. measuring weight and height) which are done every time in order to see the trend over time.  There are particular tests that should be done at specific intervals to check for particular problems.  But for this scenario it is attempting to diagnose a problem to a degree that one can distinguish between one of many possible follow on steps.  You might visualize the process as something like this:

There is a step at the beginning “Confer” where Betty talks with the patient to gather information on how the patient feels and what problems are presenting.   Then there is a step which is involves running some further tests, which might be done by the doctor or by other people who specialize in tests, like a MRI, CT Scan, or blood tests.  After than there are a large number of treatments, maybe 50 or 100 of them.  Betty will weigh the evidence and determine that a particular treatment should be tried. Betty will refer the patient to the specialist by assigning the treatment task to the referred doctor.  The system will take over from there.

Referral and Introduction

Charles is the back care specialist to whom the patent, Alex, is being referred.  Before Charles can do anything, he must be introduced to Alex, and he must be introduced to the task.  Let’s call what is happening here “Task Introduction“.

Many simplistic treatments of business process consider task introduction to be outside the scope of the business process.  Tasks are assigned the minimal of explanations: “Approve Expense Report” assumes that the performer of this task knows what needs to be done, and those additional details of what must and must not be approved is learned through a completely different channel.

Task Introduction then includes everything from when Charles first learns that there is something to be done, until when he understands what he is being asked to do.  Basic notification can be accomplished with an email message.  Charles might then follow a link to a web UI that contains additional detail about the patient, the doctor, and the test results that have already been produced.

Remote Participation

As Pictured above, Charles might interact directly with the case system of the primary care doctor. This would certainly be the case if these doctors worked for the same company, but in this scenario they don’t. Instead, Charles clones the case instance.

Case Instance Cloning

Case cloning is when a local case is made that matches the remote case, and contains a copy of the accessible contents of the remote case.  Why would Charles want to do this?  Because Charles does not have rights on the remote case to make changes there.  If he has his own goals he wants to perform, such as “research” and “recommend” as shown,then he needs a place to manage those goals.

The clone contains a copy of the data, like the CT Scans or MRI scans, but not because Charles needs access.  Charles can access the originals, but if Charles has others in the office, then he needs a copy so that he can allow his assistants to access and process them.  Goals can be set in this cloned case, more documents and information can be stored and managed there.  It is even possible that if updates were made

Where does the Personal Assistant come in?

The personal assistant can automatically accomplish some of this.  Remember that Betty assigns a task to Charles as a way of referring this work to him.  Instead of picking this up manually, it might be picked up by a personal assistant, which then uses rules to do a lot for Charles before he has to get involved.  Here is what the personal assistant software can accomplish:

• Receiving and screening notification – if Charles gets a lot of requests to do task, those emails can go to the personal assistant which can filter for interesting tasks.
• Task Introduction – pick up additional information about the task to evaluate using a set of rules whether this task is interesting.
• Task Acceptance – sending a notice back to the sender that the offer is interesting and going to be considered by a human.
• Clone Project – based again on rules it may automatically retrieve all the
• Determine the Right Template – again based on rules, and start the process if necessary.
• (& Transform – I will talk about this in the next post)
• (& Synchronize – when necessary)

Is that all?  Why give this the trumped up name of personal assistant?  The reason is that this piece of software is “acting on Charles’ behalf” which is what personal assistants do.  The task was assigned to Charles, but the personal assistant picked it up.  The personal assistant might actually do some negotiation, perhaps clarifying the terms.  The personal assistant is bridging from the Charles’ environment, working for Charles, and reaching over to act on Betty’s environment.  The personal assistant might react by directly taking action on the remote system based on rules.  Or it might as pictured do some ground work for Charles by setting up an environment for Charles to complete the work in his own way.  This is precisely where you need personal assistants to take action.

Isn’t this just a Subprocess?

It would just be a subprocess if it was designed from the beginning to fit together by a single designer.  In the next post I will talk about the problems that arise because the system that Betty uses is designed by one company, and the system that Charles uses is designed by someone else.  The personal assistant has a critical role in getting these to work together.