More than 100 kinds of Excel documents, built in a very short time Development is now far more efficient than before
Left: Yukei Wachi, Leader, New Core System Development Team, Information Systems Department
Right: Toru Kato, New Core System Development Team, Information Systems Department
Overview
Happinet Corporation adopted Docurain, the cloud-based document generation engine from Docurain Inc., to meet its sales department's needs for document output. In system development, implementing document output comes with costs that can't be ignored. Docurain simplified that work, and Happinet now generates more than 100 kinds of documents.
We asked Yukei Wachi, who led the adoption of Docurain, how the company came to adopt it, what it uses it for, and what potential he sees in Docurain going forward.
About the company
Happinet Corporation, which has marked its 50th anniversary, is a leading intermediate distributor of toys. A member of the Bandai Namco Group, it is a full-line entertainment trading company that handles not only toys but also visual and music products, video games, and amusement products.
Interviewee
Yukei Wachi, Leader, New Core System Development Team, Information Systems Department
Wachi has been with Happinet Corporation's Information Systems Department since October 2015. Over his career, he has worked on standardizing development processes, designing architecture, and managing large-scale system development, among other things.
Wachi has translated many acclaimed technical books into Japanese, including Eric Evans's Domain-Driven Design, a book on software design methods. He also gives talks regularly at conferences and other events.
Background
With the core system being replaced, document output had to be implemented in a short time
Besides toys, we run several other businesses, including video games, visual and music products, and amusement, and we've been running a core system that we developed in-house. But this core system is old and hard to maintain, so with the next 10 years in mind, we've started a project to replace it.
People in our business departments have used the current core system for many years and are used to it, so the new core system needs to carry over its usability to some extent. One such feature is exporting various kinds of data.
Selection process
For users working with data, Excel documents work best
When we set out to implement the export feature, we considered which format would be best, and in the end we decided Excel documents were the best choice.
That's because the documents we mean here weren't mainly for reporting or submitting to business partners or others outside the company. They were mainly for internal use, by individual operators or by teams.
In practice, these were cases where people wanted to save or print what they saw on screen: lists of expenses paid, product inventory lists, lists of how that inventory is allocated, and so on.
Another reason is that our work involves heavy use of Excel every day, so people in the business departments are highly proficient with it and readily accept documents in Excel. Excel documents are also more than just output: people can easily work further with the data in them, which we expect to raise productivity further.
As for how to implement it, since these were list-style Excel documents, Apache POI, a well-known open-source library for working with Excel files, was a strong candidate.
Results
Choosing Docurain over an in-house framework greatly reduced initial development effort
If you actually try to implement Excel document output using Apache POI directly, you need a fairly deep understanding of the Java API, the POI API, and the Excel specifications, so it isn't easy. And beyond development effort, we also had performance requirements, such as generating a sheet with dozens of columns and thousands of rows within 10 seconds.
Also, although most of the documents we output are lists, there are more than 100 kinds of Excel documents to cover. Developing them one by one would never have met the schedule, so we needed an efficient way to develop them.
We were looking for a product that could solve these problems, while also considering building our own framework to reduce what the developers of each feature had to keep in mind. That's when Docurain caught our attention.
I'd say Docurain's biggest appeal is its simplicity. You create a template in Excel, prepare simple JSON as the input data, and connect through a simple REST API, and that's all it takes to start generating all kinds of Excel documents right away. It also fully meets our performance requirements.
Creating Excel templates requires no programming knowledge, just ordinary Excel skills. We've drastically cut the development costs that an implementation using Apache POI would likely have incurred, and we're developing Excel documents quickly.
With Docurain, we expect to be able to respond quickly to the many Excel document requests that will likely keep coming from staff within the company. So it's an excellent choice not only for initial development costs but also for operation and maintenance costs.
Finally, we asked about his expectations for Docurain
Because the things you have to think about when developing Excel documents are narrowed down to the bare minimum, I think business departments may be able to take an active part in planning documents.
With conventional document development, when a document layout needs to change, the process is to send a request to the IT department and work out the details with them.
With Docurain, on the other hand, in some cases all it takes is changing the Excel template. And if you can write Excel macros, you can create even more elaborate documents.
You could say this ties into the idea of having business departments, rather than the IT department, lead the development of business systems. And when you're trying to promote what's called end-user computing, I'd say it can be a good tool.