Members who know the work build apps using generative AI

SYSLABO has been working on citizen development, in which members who carry out the work take part in application development, using ServiceNow App Engine and Now Assist for Creator.

Members who do the work every day know well what information they need, when decisions are made, and the situations their usual procedures cannot handle. If that knowledge is reflected in building apps, it becomes easier to design systems that fit the actual work.

Through a review of our internal systems, we created opportunities for business members to learn, build prototypes, and think about improvements. Bringing in development support from generative AI, we are putting into practice a way to connect knowledge of the work with system development.

• Reviewing, as a whole, the systems that supported individual work

  Behind this effort was the aging of internal systems we had used for many years. Another challenge was that separate apps had been built at each site to make work easier, which made it hard to see the overall picture of our systems and how they were managed.

  In particular, the system supporting sales activities and the sales management system were separate, and information about prospects before an order was not connected with information about sales after the order. We therefore adopted App Engine as the platform for reviewing both areas of work as one.

  When business systems are separate, even if each person enters the information they need, it may have to be copied and checked before it can be used in the next task. For example, if it is unclear which information to update and whom to tell when the outlook for a deal changes, the effort of checking the latest status also increases.

  In this kind of review, it is important to understand the role the existing systems have played. By checking why certain information is entered, whose decisions it is used for, and what exceptions it handles, we can think about which functions to keep and which parts to change.

  We believe that organizing how the work connects, and building a system in which information is passed along naturally, is essential when reviewing systems.

• Members who know the work take part in development

  Our project was carried out by eight members, mainly from sales. In addition to study sessions for the members, we held build sessions to create prototypes called mockups, and considered concrete applications while learning.

  The value of business members taking part in development is that they can bring the decisions they make in their work into the design.

  For example, in an app for checking the outlook of deals, a list of deal names and amounts is not always enough. The information needed and how it is used change depending on when the sales are expected, at what stage the information is updated, and with whom the reasons for a changed judgment are shared.

  Such conditions may seem natural to the people who do the work every day, but they are not always conveyed to the people who develop the system. When members who know the work take part in prototyping, it becomes easier to confirm concretely how a system will be used and the reasons behind decisions that are hard to convey in words alone.

  Citizen development is also an opportunity to use the knowledge and insights gained in daily work to improve that work.

• Turning ideas into prototypes with generative AI

  Supporting this effort is Now Assist for Creator, a development support feature that uses generative AI. We also used App Generation, which generates application mockups.

  When people only describe in words what they want to build, the image of the finished product may differ between those who use it and those who develop it. With a prototype, they can check those differences while looking at the same screens and behavior.

  For example, they can check whether the order in which information is reviewed matches the actual work, whether any information needed for decisions is missing, and whether handing work over to another person will cause problems. Using a prototype can also reveal requests that did not come up in the first explanation, or steps that would be better reviewed.

  The first draft prepared by generative AI is a starting point for this kind of discussion. Its value lies in letting business members give shape to their ideas, check them with the people around them, and move on to the next revision.

• Generative AI that makes prototyping and learning easier

  The ServiceNow official case study on our use of Now Assist for Creator introduces our sense that the effort to create mockups can be reduced by more than 50%, and an estimate that the speed of App Engine development and skill acquisition doubles.

  When prototyping takes less time, there are more chances to test ideas. Instead of building something once and stopping there, we can reflect what we notice in use, try other approaches, and look for a form that fits the work.

  For learning as well, it is meaningful to be able to think while looking at a concrete app. When what you want to build is linked with how to use the features, it becomes easier to see what you need to learn.

  In citizen development with generative AI, members can deepen their understanding by moving back and forth between prototyping and learning. Even when something is unclear, they can ask questions based on the prototype or consult experienced members, which makes the next thing to check concrete.

• Turning users’ insights into continuous improvement

  When business members take part in development, it becomes easier to turn the inconveniences they feel while using a system into discussions about improvement.

  For example, if the same information is checked repeatedly, we can consider whether there is a problem with how the information is shown, whether the timing of updates does not match, or whether roles between people are unclear. Small daily inconveniences become a chance to review how the work itself is done.

  This requires a place where insights can be shared, and someone to check them with. Being able to talk while looking at a prototype or an actual app helps business members and development members align their understanding.

  We believe that repeating this kind of dialogue leads to an environment where it is easy to take part in improvement. Seeing one’s own opinions reflected in a system also becomes a trigger for thinking about the next improvement.

• Business members and technical specialists bring their knowledge together

  To connect citizen development with actual work, it is also important to have a structure in which business members and technical specialists can work together.

  Business members can make concrete the information needed, how it is used, and the conditions for decisions. Based on these, technical specialists consider how data is held, integration between systems, access rights, performance, testing, and more. By bringing together both kinds of knowledge, we can design systems that are easy to use and can be operated with confidence.

  To keep improving after people change roles, it is important to be able to share why a specification was chosen and which changes will affect other work. Recording the knowledge gained during development and the reasons for decisions is also a foundation for continuing in house development.

  Precisely because generative AI has made prototyping easier, we value checking what has been built from both the business and technical sides, and shaping it into something that can be used for the long term.

• Applying our own practice to supporting clients’ in house development

  SYSLABO proposes and implements ServiceNow for clients, and at the same time is a user that applies it in its own work.

  Learning development while carrying out daily work, using prototypes to make requests concrete, and sharing ideas among members with different expertise all involve difficulties that only become clear by actually doing them.

  When supporting clients’ in house development, we also need to consider, based on each client’s situation, which work is easiest to start with, how business members will take part, and where technical specialists should provide support.

  We will apply the experience gained through our own practice to discussions with clients and to development support. We provide support with learning, prototyping, implementation, and operation in view, so that the insights of people who know the work well can be turned into concrete improvements with the help of generative AI and technical specialists.