Automating Internal Inquiries with AI
In an industry where rapid inquiry handling is crucial,
Comix Co., headquartered in Shibuya, Tokyo, has taken a significant step toward efficiency. They have implemented a half-automated system using generative AI to respond to internal inquiries concerning guarantee products. This pioneering effort is part of their broader initiative to embrace digital transformation and optimize their operations.
Background and Challenges
The guarantee services sector includes various products such as rental debt guarantees, medical expense guarantees, and care cost guarantees. The increasing number of products, coupled with frequent policy amendments, has led to a surge in internal inquiries—up to
700 to 800 inquiries monthly. Previously, the staff manually handled each inquiry, which imposed a heavy workload on various departments.
Recognizing that historical inquiry and response data was already available internally, Comix identified an opportunity: "The answers are already within the company; can we let AI extract and present them?" This realization sparked their initiative to integrate AI into their operations.
Solution Overview
Comix's solution consists of multiple components aimed at streamlining inquiry responses:
1.
Harnessing Historical Data: The existing Q&A data was leveraged to create a streamlined database via spreadsheets, forming a solid foundation for automated responses.
2.
AI Response Generation: The system integrates Claude Code with Google Apps Script (GAS) to automatically ingest inquiries received via email and generate response proposals based on accumulated data.
3.
Human Oversight: Although the AI generates suggestions, human personnel are responsible for validating and finalizing responses, ensuring that quality and accuracy are maintained. This approach emphasizes a hybrid model where human judgment complements AI's efficiency.
This gradual approach towards automation allows for continuous assessment of accuracy and trustworthiness before expanding the AI's role.
CIO Involvement and Hands-On Learning
A distinctive aspect of this initiative is the direct involvement of the Chief Information Officer (CIO), who has dedicated time to learn and apply Claude Code. This included:
- - Initial workshops to assess operational challenges and prioritize inquiry automation.
- - Hands-on sessions where the CIO built the Claude Code environment, experiencing firsthand the ability to create automated processes from scratch.
- - Regular meetings where discussions with the AI led to the creation of design documents that directly informed the implementation of new systems during live sessions.
This proactive learning environment promotes a culture of collaboration between AI and human operators, equipping the entire organization with the skills necessary for effective AI deployment.
Unique Features and Advantages
Comix's approach emphasizes several key advantages:
- - Design-First Approach: Reducing errors by allowing human operators to finalize specifications before AI implementation.
- - AI-Driven Implementation: Once specifications are clear, AI executes tasks without needing further input, leading to rapid deployment.
- - Real-time Adjustments: Minor adjustments can be communicated through natural language prompts, streamlining user interface modifications.
This efficient division of labor minimizes rework and maximizes the value derived from AI capabilities, allowing even non-engineers to produce practical tools swiftly.
Target Users and Application Scenarios
Comix's services are particularly beneficial for:
- - Executives managing businesses centered on guarantee services and diverse products.
- - Operational managers overwhelmed with inquiry handling.
- - IT departments aiming to advance the application of generative AI within their organizations.
Practical use cases include:
- - Administrative tasks like handling internal FAQs and processing various requests.
- - Addressing inquiries related to product specifications in real-time.
- - Enhancing organizational strength by documenting and utilizing existing Q&A resources and manuals.
Future Prospects
Comix aims to further enhance its system for semi-automated inquiry responses by integrating educational resources for company-wide utilization. Future plans also include:
- - Centralized management of guarantee product data, creating a single source of truth for information and workflows.
- - Developing customer-facing solutions based on the refined internal processes, potentially revolutionizing their service delivery.
By linking AI implementation directly to tangible business processes, Comix is committed to improving the quality and speed of its guarantee services while transforming its operational landscape, leading the charge into the future of automated services.
For more inquiries regarding AI utilization support, please reach out through the contact details provided.