PERSOL BUSINESS PROCESS DESIGN Conducts Proof-of-Concept in Setagaya Ward on Knowledge Generation Using Call Recording Data from Telephone Service Desks
PERSOL BUSINESS PROCESS DESIGN has completed a proof-of-concept in Setagaya Ward, Tokyo, to verify the creation of a knowledge base for ensuring smooth call transfers and providing accurate information to residents. The experiment involved collecting and analyzing recorded data from inquiries made by residents to the main office telephone line, visualizing response know-how based on staff experience, and structuring the knowledge for future AI-powered automated responses.
📋 Article Processing Timeline
- 📰 Published: April 3, 2026 at 20:00
### PERSOL BUSINESS PROCESS DESIGN Conducts Proof-of-Concept in Setagaya Ward on Knowledge Generation Using Call Recording Data from Telephone Service Desks
**~ Verifying Knowledge Creation to Build a System for Smoothly Transferring Calls and Accurately Delivering Information to Residents ~**
PERSOL BUSINESS PROCESS DESIGN CO., LTD. (Headquarters: Minato-ku, Tokyo; President & CEO: Kazuyuki Ichimura), a part of the PERSOL Group which operates under the vision 'Work, and Smile,' has completed a 'proof-of-concept experiment on knowledge generation using call recording data from telephone service desks' in Setagaya Ward, Tokyo (Mayor: Nobuto Hosaka). In this experiment, recorded data of responses to inquiries from residents to the main office telephone line was collected and analyzed. The goal was to visualize the response know-how, which relies on the experience and judgment of staff, to ensure accurate call transfers, and to structure this knowledge in anticipation of future AI-powered automated responses.
### ■ Background
In recent years, as residents' lifestyles and administrative services have diversified, inquiries to local governments have become more complex. In Setagaya Ward, while contact information for each department is available on the ward's website, some residents, unsure of where to direct their inquiries, caused a concentration of calls to the main office telephone line. At the main office, staff handle these calls concurrently with their regular duties, creating a burden and a need for improved efficiency.
Furthermore, due to the wide variety of inquiries, accurately grasping the issue and smoothly connecting the caller to the appropriate department within a limited time has largely depended on the individual experience and knowledge of each staff member. This reliance on individual skills, which often remains as tacit knowledge and is not sufficiently accumulated as data, was also a challenge.
Against this backdrop, aiming to build a system that can provide information to residents more clearly and accurately without making them wait, the Setagaya Ward DX Promotion Department and PERSOL BUSINESS PROCESS DESIGN undertook this proof-of-concept experiment to datafy inquiry trends and consider improvements to reduce the burden on staff.
### ■ Experiment Overview and Results
Conducted from January to February 2026, this experiment involved collecting and transcribing approximately 1,800 call recording data from staff responses at the main office telephone line. The content was then structured, organized, and databased as knowledge. As a result, the accumulated knowledge (approximately 270 items) was confirmed to contribute to identifying the purpose of the inquiry and facilitating smooth transfers to the appropriate department. This is expected to be utilized for delivering information to residents accurately and without delay, and for verifying the feasibility of future AI-powered automated responses.
Additionally, the analysis visualized the process of identifying the purpose from diverse inquiries, revealing clear trends of routine inquiries concentrated during specific periods, such as those related to My Number Cards and tax returns. This confirmed that many issues could be resolved without staff intervention, paving a concrete path toward the future implementation of AI-powered automated responses and chatbots, and demonstrating a way to reduce the burden on staff.
### ■ PERSOL BUSINESS PROCESS DESIGN's Strength in Knowledge Creation
In creating a knowledge base for inquiry responses, it's not always a one-to-one match between an inquiry and a piece of knowledge. Operators respond from various angles to connect the inquirer with the knowledge they truly need from among many inquiry patterns. Our strength lies in our ability to build a database that includes this operator-held knowledge.
### ■ Future Outlook
Based on the results of this experiment, PERSOL BUSINESS PROCESS DESIGN will proceed with building a model for standardizing and improving the efficiency of main telephone line responses, and will organize tasks into those that require human intervention and those that can be resolved by residents themselves through FAQs and automated responses. Furthermore, we will support the expansion of this initiative to other telephone service desks within the ward office and contribute to the development of mid- to long-term administrative DX measures such as the introduction of AI-powered automated responses and chatbots. We will also leverage the insights gained from this experiment to support the efficiency and quality improvement of service desks in municipalities nationwide.
**~ Verifying Knowledge Creation to Build a System for Smoothly Transferring Calls and Accurately Delivering Information to Residents ~**
PERSOL BUSINESS PROCESS DESIGN CO., LTD. (Headquarters: Minato-ku, Tokyo; President & CEO: Kazuyuki Ichimura), a part of the PERSOL Group which operates under the vision 'Work, and Smile,' has completed a 'proof-of-concept experiment on knowledge generation using call recording data from telephone service desks' in Setagaya Ward, Tokyo (Mayor: Nobuto Hosaka). In this experiment, recorded data of responses to inquiries from residents to the main office telephone line was collected and analyzed. The goal was to visualize the response know-how, which relies on the experience and judgment of staff, to ensure accurate call transfers, and to structure this knowledge in anticipation of future AI-powered automated responses.
### ■ Background
In recent years, as residents' lifestyles and administrative services have diversified, inquiries to local governments have become more complex. In Setagaya Ward, while contact information for each department is available on the ward's website, some residents, unsure of where to direct their inquiries, caused a concentration of calls to the main office telephone line. At the main office, staff handle these calls concurrently with their regular duties, creating a burden and a need for improved efficiency.
Furthermore, due to the wide variety of inquiries, accurately grasping the issue and smoothly connecting the caller to the appropriate department within a limited time has largely depended on the individual experience and knowledge of each staff member. This reliance on individual skills, which often remains as tacit knowledge and is not sufficiently accumulated as data, was also a challenge.
Against this backdrop, aiming to build a system that can provide information to residents more clearly and accurately without making them wait, the Setagaya Ward DX Promotion Department and PERSOL BUSINESS PROCESS DESIGN undertook this proof-of-concept experiment to datafy inquiry trends and consider improvements to reduce the burden on staff.
### ■ Experiment Overview and Results
Conducted from January to February 2026, this experiment involved collecting and transcribing approximately 1,800 call recording data from staff responses at the main office telephone line. The content was then structured, organized, and databased as knowledge. As a result, the accumulated knowledge (approximately 270 items) was confirmed to contribute to identifying the purpose of the inquiry and facilitating smooth transfers to the appropriate department. This is expected to be utilized for delivering information to residents accurately and without delay, and for verifying the feasibility of future AI-powered automated responses.
Additionally, the analysis visualized the process of identifying the purpose from diverse inquiries, revealing clear trends of routine inquiries concentrated during specific periods, such as those related to My Number Cards and tax returns. This confirmed that many issues could be resolved without staff intervention, paving a concrete path toward the future implementation of AI-powered automated responses and chatbots, and demonstrating a way to reduce the burden on staff.
### ■ PERSOL BUSINESS PROCESS DESIGN's Strength in Knowledge Creation
In creating a knowledge base for inquiry responses, it's not always a one-to-one match between an inquiry and a piece of knowledge. Operators respond from various angles to connect the inquirer with the knowledge they truly need from among many inquiry patterns. Our strength lies in our ability to build a database that includes this operator-held knowledge.
### ■ Future Outlook
Based on the results of this experiment, PERSOL BUSINESS PROCESS DESIGN will proceed with building a model for standardizing and improving the efficiency of main telephone line responses, and will organize tasks into those that require human intervention and those that can be resolved by residents themselves through FAQs and automated responses. Furthermore, we will support the expansion of this initiative to other telephone service desks within the ward office and contribute to the development of mid- to long-term administrative DX measures such as the introduction of AI-powered automated responses and chatbots. We will also leverage the insights gained from this experiment to support the efficiency and quality improvement of service desks in municipalities nationwide.
FAQ
What was the purpose of this proof-of-concept experiment?
To analyze and convert call responses, which depended on staff experience, into a knowledge base using AI, thereby making resident services more efficient and standardized, while also building a foundation for future automated response systems.
What were the specific results?
Approximately 270 knowledge items were generated from about 1,800 calls. It visualized trends in routine inquiries, such as those related to My Number cards, and identified tasks that could be automated by AI.
What are the future prospects?
The plan is to expand this model to other municipalities to support the efficiency of government service counters nationwide. They will also proceed with implementing tools like AI chatbots.