Challenge
The breeding data at the National Institute of Fisheries Science had no standardized format. Each researcher organized their data in Excel in their own style, so even the same kind of data varied in structure and fields from one person to the next. Data organized this way is hard to use directly when it later needs to be analyzed or extended with AI.
There had been earlier attempts to gather the data in one place, but they forced entry into a fixed format. Because the input method constrained how researchers worked, the systems went largely unused, and the data scattered back into individual files.
Approach
Alpaon built a central platform to bring the data together, and designed it with extensibility and flexibility as the first priority.
The key was not to impose a rigid input format on researchers. Since the reason earlier systems had failed to take hold was the enforcement of a format, we struck a balance: researchers keep their existing way of working and their freedom to record, while the data still accumulates within a system usable for later analysis and extension.
At the same time, we designed the structure so the accumulated data could be extended to many uses later. We left room to integrate real-time sensor data and to let AI models draw on the accumulated data directly.
Outcome
The National Institute of Fisheries Science gained a foundation for accumulating its previously scattered breeding data on a single platform. Because researchers' freedom to record was preserved, the platform can actually be adopted in practice, and with the data gathered in a consistent system, a base for later analysis and application is in place.
Real-time sensor integration and AI model use are not yet built, but the platform is designed to accommodate them, so it can be extended later without rebuilding.