Hi, there!

In the article “What is the Value that the Parallelization of Workflow Brings?“,I mentioned that large effects can be obtained by arranging parallel processing in the Workflow.

In this article, I will introduce a case study which had great results by incorporating “parallel processing” into an existing Workflow.

Parallel processing shortened the period of time for evaluation work to 1/5

In a certain mail-order company, they have succeeded in reducing the period of time for evaluation work in the workflow of private label product development to 1/5 by incorporating “parallel processing”. (The name of the mail-order company is not disclosed)

In merchandise development of private brands, the “evaluation” work related to purchasing is an important task. There are three evaluation tasks: evaluation of suppliers, evaluation of manufacturing / production systems of products to be purchased, evaluation of the logistics system concerning the goods.

These evaluation tasks were done one after another in the past. That meant, after the evaluation of the supplier was done, they evaluated the manufacturing and production system, then evaluated the logistics system after the evaluation of the manufacturing/production system was done, and so on.

At the place of goods planning, they wanted to shorten the time required from “drafting” to “completion” (lead time), so that they could proceed with the planning of more goods. Thereupon, they considered parallelizing these three evaluation works.

I suppose that they had to proceed with the evaluation work one after another since they used to manage each goods planning project with documents in paper. By using a workflow system instead of doing it with paper, it became possible to reliably manage the situation of the projects, even when processing parallelly which is said to be difficult to manage.

They say that the time from drafting to completion has been shortened to one-fifth since they made it into parallel evaluation.

Sooner or later, I will describe the business management including parallel processing using the workflow system.

That’s it, for today!

Hi, there!

If you can find a part where you can parallelize in a Workflow, your productivity will be improved greatly. In this article, I am going to talk about “parallel” processing in a Workflow.

What is “parallel” processing in a Workflow?

A Workflow is a series of Steps (tasks) necessary to produce some results, such as making products and providing services, arranged in a proper order and connecting them with arrows.

There are two ways to connect these Steps in Workflow diagram, “serially” and “parallelly”.

To define “serial” versus “parallel”, I would like you to imagine the concept of “balls of work”. A “work ball” is what shows which process (Step) in the workflow is being handled and what is being processed in it.

When connecting “serially”, only one ball is at a given Step (the Step that is being processed).

Whereas, when connecting “parallelly”, several balls can be at different Steps as well. In “parallel” processing, at the beginning when there was only one ball, it can become two or more at a certain point.

Advantage of “parallel”processing

If you are looking at a workflow diagram and you can find a part where you can process “parallelly”, it is a great opportunity to improve the business process.

If you can connect multiple Steps in parallel and process them simultaneously, you can reduce the time from the start to the end of the entire flow. That is the biggest advantage of “parallel” processing.

Assuming that the time required for processing Step A as t1 and the time for processing Step B as t2, the time needed to complete these two Steps is the sum of t1 and t2 if Step A and Step B were connected in series. Whereas, if Step A and Step B are connected in parallel, the time to complete these two steps will be only the one that takes the longest, either t1 or t2.

“t1 + t2 > t1 or t2”, thus serial is greater than parallel, it is obvious that parallel processing can shorten the processing time required for the entire workflow.

Since there is such an advantage in parallel processing, I think that it is worthwhile to consider whether or not there is a part in the process which can be done “in parallel” in the Workflow diagram, including the review of the entire work.

Is “parallel” processing difficult to manage?

By arranging “parallel” processing in the Workflow, you will be able to gain the benefits described in the previous section. However, adding “parallel” processing makes management more difficult.

Since the original, single work ball becomes several, it becomes difficult to manage where each of the balls are.

The management of parallel processing will become easy by using a Workflow system (BPM system). Also, there are cases where business improvement was achieved by arranging parallel processing. I would like to continue talking about parallel processing in the future.

That’s it, for today!

It will take five more years for bank APIs to become usable…So…
“Invoice + bank remittance” is …, after all, out of date now…

“When you mail a billing invoice in paper, it will be paid until its deadline. Indeed, I believed it was the way to do it until a couple of years ago, but now, it is old-fashioned.

1. Bill invoice in paper

For example, when you mailed a couple of invoices, naturally it will be paid combined. Then a person would identify the payments…really analog!

Speaking of human tasks, it reminds me an old time when salaries were paid in brown envelopes, even to the pennies.

2. Online banking

People are talking about ‘Fintech’, yet it is useless.

Even in “Bank API as a national policy”, only “Reference type APIs” have been released partially after a long way. Moreover, the Nationwide Bank Data Communication System in Japan has its unique problem in text data communication. That must be corrected…

3. Full automation until settlement

I would like to process from billing to payment settlement without human intervention.

That’s right! “The ones born in the 70’s” should be the ones to say good-by to “bad ways of doing things” done so far. (?)

If Banks are not yet usable, let’s use a “Settlement agent” (Money transfer agent)!

1. Register “electronic invoice” on the settlement agent,
2. Make request to the settlement agent and have them email the “customer”,
3. Receive a report if there was a deposit from “customer” to the settlement agent.

In this way, the workflow would become unmanned.

4. Correspondence with Technical Support

Conversations with Technical Support of a money transfer agent regarding APIs… (This is fiction!)

Me: “Hey, hey, even if I specify the template ID it is not reflected in the invoice, right?”

(1 month later)

Support: “Ah, that’s true. I confirmed with our developers, they said the specification is not to reflect it. Our document is wrong.” (What?)

Me: “And also, even if I have set JST as timezone for date of invoices, it doesn’t seem to be reflected.”」

(1 month later)

Support: “That is the specification. Even if you specify the timezone, it is shown in PST.” (Huh?)

Me: “Also, the phone number for the billing source is not indicated on the bill…”

(1 month later)

Support; “Ah, that is a mistake in the document. The Phone object works if requested as child object of address object. Even though it is described as a sibling in the document. (I’ll kill you!)

note) In the capture image of sample code below, relationship of address and phone is wrong.

Well, the APIs designed by the settlement agent (money transfer agent) is really rational. And regarding the “REST Invoicing API” of “PayPAL Inc.”, I think it is awesome. (For overseas remittances, banks cannot match it.)

(But, there are some problems, at the same time… Especially, Support…)

5. Practical example

Incidentally, in an open event “Super Networking 2017” (2017-06-17), for which I am involved in the committee, electronic invoices (PayPal invoice) are practically used for collection of participation fee.

If you are up for the challenge, please join. (!?)

An email of “electronic invoice” will be delivered immediately… supposedly…

Case of accessing Questetra’s API from the midstream of Questetra’s process.
Happy New Year. Thank you for your continuous support this year.

Though I have written blog posts about various system collaboration cases so far, I would like to write more of those kind of posts this year as well. Besides, I personally like the topic of system collaboration because it can extend options in many aspects.

In this post, I would like to introduce a method for accessing Questetra’s API from a Process of Questetra.

1: Use Case

As far as I know, the method has been used in the following cases.

• After approving User addition/deletion request, add/delete User account of Questetra. (API for addition/deletion of User)
• To launch the Task regularly in order to have a User changing the password periodically, and reflects the changes to the User settings of Questetra. (API for addition/deletion of User)
• Change the organization setting of Questera so that the User who makes approval is changed for a certain period, and reverse the change after reaching a deadline. (API for adding/deleting user’s affiliation to organizations）
• In Approval flows such as approval request or decision-making request, obtaining the organizational information of the applicant User and determine an approver on the request by the computing, and then allocate the Task of approval to the User. (API for querying settings of User, Organization, and Role)
• Calculate the progress rate and delay degree by referring to the data of certain Process that performs process management and write the result to the original Process. (API for acquiring processing result of business flow, API for executing task)

2: General classification of Questetra’s API

If I roughly classify the APIs of Questetra, it will be;

• A. To prepare by placing as parts in the Workflow: Operates such tasks as sending and receiving of HTTP request etc. in the midstream of a Process.
• B. Other than the above: Operate such tasks as referring and updating of Users or data in a Process that Questetra retains.

The latter is a REST API, which can be accessed with “Throwing Message Intermediate Event (HTTP)” which is one of the former.
However, even if from the same Questetra environment, it requires authentication when accessing the API of the latter.
Besides, in the case of processing (computing, etc.) based on the obtained data, you need to parse the response with “Script Task”.

• For “Throwing Message Intermediate Event (HTTP)”, please refer to the following.
  　M225: Auto Sending of HTTP Request with Business data Inserted
• For “Script Task”, please refer to the following.
  　M230: Auto Executing Complicated Data Processing (ECMAScript)

3: Details of settings

In this post, I will describe the case of accessing an API for adding Users through Basic authentication. (A simple case that does not require response parsing with “Script Task”.)

The following is a sample Workflow App in which automatically adding a User account after a request has been approved by superiors.

The specification of User addition API is described in the API reference.

*The API manual can be displayed from the menu on the left side of each Questetra environment, and the URL is as follows.
https://xxxxx.questetra.net/Swagger/index

The link to the manual of online demo environment is as follows. There, you can see the manual immediately even if you do not have a Questetra environment.
System Settings API
Workflow API

The following is a screenshot for of the corresponding API.
* In order to perform user addition processing, authentication by a User with System Administrator Authorization is needed. As it might be naturally expected.

Based on the API specification above, the settings of “Throwing Message Intermediate Event (HTTP)” can be configured as follows.

• Specify the URL of API you want to access as Access URL (If in the same Questetra environment, you can also specify a variable such as “${var[applicationRoot]}”).

• Specify Username and Password for Basic Authentication (Use API Password mentioned above).

• Specify the parameters of the API to be accessed as sending parameters.

 You can download an archive of the sample Workflow App above from here (zip expands qar file).
* Change the part of Basic Authentication according to your environment.

For the other API specification of Questetra, please refer to the following.

• Related manual: M317: Controlling OAuth2 Authorization Access and Basic Authentication Access from External
• Related manual: M318: Controlling Accesses from External to [Message Catch Event API]

4: Closing

By utilizing this method, we believe it is possible to automate and streamline more advanced business will flow.

If you have any questions about this post, please feel free to contact us via Contact form.

Apply for Starter Plan (Free) Here

By signing up for your Free account of your own Questetra, you will be able to use all the Questetra features including what I mentioned above.