Generative AI analysis provides insights in just minutes, promoting evidence-based policymaking (EBPM) in local governments
To media representatives:
November 5, 2025
Saison Technology Co., Ltd.
Saison Technology Co., Ltd. (Head office: Minato-ku, Tokyo; Representative Director, President and CEO: Makoto Hayama; hereinafter referred to as Saison Technology) is pleased to announce that, with the full cooperation of Meguro Ward and the support of the Japan Agency for Local Public Entity Information Systems (hereinafter referred to as J-LIS), it has jointly conducted a demonstration experiment to link various data held by local governments and use generative AI to evaluate and analyze policy planning.
This demonstration experiment demonstrated that insights (suggestions) for policy planning using local government data and generative AI could be obtained in just a few minutes from prompt instructions, demonstrating that local governments will be able to promote evidence-based policy making (EBPM) *1 in the future, replacing the previous manual data analysis and evaluation methods.
background
In order for local governments to provide and manage public services that reflect the will of local residents, solve local issues, and improve the lives of residents, it is important for them to evaluate policies based on evidence using the data they possess.However, the current situation is that many local governments are unable to fully utilize the vast amount of data they possess to evaluate policy plans due to their busy daily work, lack of dedicated staff, and limited budget execution.
To solve these issues, Saison Technology, with the full cooperation of Meguro Ward and the support of J-LIS, conducted a demonstration experiment to link various data held by local governments, input it into generative AI for analysis, and evaluate policies based on insights such as extracting trends.
Overview of evaluation demonstration using data held by local governments and generated AI
In this demonstration experiment, data on facility usage and costs, primarily from open data held by Meguro Ward, was provided, and under the theme of "effective utilization of ward-owned assets," generative AI was used to verify whether ward facilities were being utilized effectively, and whether data analysis by generative AI could reject the premise that "a major review of the business is necessary."
Saison Technology used the iPaaS "HULFT Square" to link multiple data sources, dynamically extracting and aggregating data from local governments in line with analytical objectives, then linked it to a closed environment generation AI to develop and build a system that verbalizes the fluctuations and correlations that can be read from the data. In addition, with the expectation of using production data in the future, J-LIS has provided them with knowledge on how to safely link and utilize data held by various local governments, including personal information.
Meguro Ward utilized the wide-ranging knowledge it had gained through its daily work to provide feedback on the validity and effectiveness of the policy evaluations output by the generative AI, thereby improving data processing and insight reports, as well as revising and improving the wording to suit local government operations.
Through this demonstration experiment, it was confirmed that by linking various data held by local governments and utilizing generation AI, the following points are useful in the policy evaluation work of local governments.
- Aggregate municipal data and visualize trends in just a few minutes
Policy evaluation work requires the collection, processing, and aggregation of various related data, followed by multifaceted analysis using data analysis tools to gain insights. Taking the example of an evaluation analysis of elderly care facilities and conference rooms in a demonstration experiment, analysis results from the perspectives of cost, finance, and human resources were obtained in just 3 to 5 minutes after prompting the generation AI. This complicated and specialized work, previously performed by local government officials, can now be carried out efficiently through data integration and generation AI. - Generate and verify numerous pattern comparisons that would be impossible for a human to do
To obtain high-resolution suggestions from data, it is important to compare data from specific perspectives, such as comparing data with other data or cities with other data. Generative AI scrutinizes valid comparison patterns and verifies numerous patterns that cannot be handled by human operators, enabling suggestions derived from analysis from various angles.
In the following example insight report, the topic is "elderly care facilities," and based on the hypothesis that major facility restructuring is necessary, the AI is prompted to perform a quantitative analysis based on data, propose necessary measures, and provide counterarguments to the proposal.
In addition to quantitative analysis of usage, facility condition, and maintenance costs, the report also proposes medium- to long-term measures such as "redeveloping the facility as a multi-generational exchange hub" and "adding childcare support and local community functions," as well as insights into staffing, such as "deploying specialized staff in health promotion and care prevention." In response to counterarguments to the proposals, the report also provides evidence that "the aim is to provide effective and efficient services rather than completely abolishing the facility" and "the quality and diversity of services will improve by making the facility more multifunctional."
- 1 EBPM (Evidence-Based Policy Making): Policy planning should not rely on ad hoc anecdotes, but should be based on rational evidence after clarifying policy objectives (from "Cabinet Office's Initiatives for EBPM")
- 2 Example: J-LIS's "Local Government Infrastructure Cloud System"
This proof-of-concept experiment has also been conducted with Meguro Ward and several other local governments, and similar results have been confirmed. While this was verified in Saison Technology's environment, going forward, Saison Technology aims to build a generative AI solution that can be shared by multiple local governments by combining it with a secure database*2 and data integration platform (iPaaS) such as HULFT Square, which are jointly used by local governments.
We have also received comments from Meguro Ward regarding this demonstration experiment.
Comment from Daisuke Takeyama, Planning and Management Division, Meguro Ward
Meguro Ward recognizes the promotion of evidence-based policymaking as an important theme, but has not been able to fully utilize the vast amount of data available in its daily operations. Through this pilot project, we have gained new insights into multifaceted analysis and rapid insight extraction by combining open data and generative AI. Based on this insight, we will use it as a reference for future approaches to evidence-based policymaking that utilize data and technology.
- 1 EBPM (Evidence-Based Policy Making): Policy planning should not rely on ad hoc anecdotes, but should be based on rational evidence after clarifying policy objectives (from "Cabinet Office's Initiatives for EBPM")
- 2 Example: J-LIS's "Local Government Infrastructure Cloud System"
About Saison Technology
Saison Technology, a Data Integrator, has a mission to "Connect the world’s data and make it useful for everyone.," and is globally deploying data integration products and IT services that form the foundation for safety and security, providing system development and operations for a wide variety of industries, including finance and distribution. Utilizing its strengths, which have enabled it to quickly adapt to changes in the environment over many years, the company is currently focusing on expanding its cloud-based data integration platform (iPaaS), "HULFT Square," as well as strengthening its efforts to implement technologies that will pave the way for the future.
- Saison Technology:
https://www.saison-technology.com/ - HULFT Series:
https://www.saison-technology.com/service/product/lineup/
[Trademark-related]
- "HULFT" is a registered trademark or trademark of Saison Technology.
- Other company names, product names, service names, etc. are registered trademarks or trademarks of the respective companies.