SoftBank transforms call center operations with AI-driven automation
SoftBank Corp., a major Japanese telecommunications provider, has embarked on a joint project with Microsoft Japan to transform its call center operations using generative AI technologies. Announced in March 2024, the initiative aims to reduce customer wait times and standardize service across over 10,000 business processes. The company previously trialed generative AI in its call centers and found significant potential for increasing automation and enhancing customer experiences. By implementing Azure OpenAI Service and Azure AI Search, SoftBank is developing an LLM-based system to provide accurate, flexible, and context-aware support, moving beyond fixed response scripts. This system will integrate RAG (retrieval augmented generation) using SoftBank’s internal databases. The phased rollout is set to begin in July 2024, with the expectation of significant improvement in customer convenience, quicker answers to inquiries, and automation of routine tasks. The strategic alliance signals SoftBank’s commitment to digital transformation via cutting-edge AI. SoftBank’s call centers manage a wide variety of inquiries relating to mobile, fixed-line, and Internet communication services, necessitating advanced handling due to process volume and complexity. With the new LLM-based system, SoftBank aims to dynamically route and resolve inquiries by leveraging both advanced prompt engineering and its own operation-specific data, ensuring responses are both rapid and accurate. The move is part of a broader push by SoftBank Corp. to lead in Japan’s digital transformation and customer service innovation through generative AI. The project showcases a deep partnership with Microsoft Japan, leveraging Azure technologies as core enablers of scale and reliability.
- Organization
- SoftBank Corp.
- Industry
- Tech & Comms
- Location
- Japan
- Published
- April 2024
Reported outcomes
Strategic outcomes
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- SoftBank Corp.
- Provider
- Microsoft
- Maturity
- Production
- Linked source
- SoftBank News
Deployed Azure OpenAI Service and Azure AI Search to build generative AI-powered support tools
Primary read
Use case focus
Showing 3 of 3
- 1Autonomous Generative AI for Multichannel Call Center Support
- 2Retrieval-Augmented Response System for Telecommunications
- 3LLM-based Dynamic Routing and Knowledge Assistants
- Call center customer wait times were too long.
- Responses to inquiries were inconsistent due to the wide variety of business processes (>10,000) and service types offered.
- Existing systems relied primarily on fixed scripts and predetermined procedural orders, limiting flexibility and personalization.
- Routine customer support tasks were not sufficiently automated, leading to increased operator workloads.
- Need for more accurate, context-aware support drawing from a wide range of company data.
- Co-developed an LLM-based autonomous customer service system with Microsoft Japan.
- Deployed Azure OpenAI Service and Azure AI Search to build generative AI-powered support tools.
- Implemented RAG (retrieval augmented generation) by integrating company-internal databases for context-aware, accurate responses.
- Moved away from fixed support scripts to flexible, dynamic guidance tailored to the actual conversation with the customer.
- Phased rollout of the AI system starting July 2024, with continuous improvement based on feedback.
- Expected reduction in customer wait times for call center inquiries.
- Anticipated increase in response accuracy and standardization across service types and channels.
- Higher automation of routine call center processes, freeing operator resources for complex issues.
- Significant customer convenience improvements by providing rapid, contextually relevant answers.
- Positioning SoftBank as a leader in digital transformation in the telecommunications sector.
Architecture
The solution utilizes an LLM-based customer support system, developed jointly with Microsoft Japan, built on Azure OpenAI Service and Azure AI Search. This system features retrieval augmented generation (RAG) by integrating SoftBank's internal databases and prompting with large volumes of operational data. The architecture moves away from traditional fixed script models, enabling dynamic query and data retrieval during customer interactions. The interplay among Azure OpenAI Service (for LLM processing), Azure AI Search (for RAG and integration with internal data), and SoftBank’s own databases forms the backbone of this architecture.
Sources & evidence1
- Customer explicitly identified
- Deployment status explicitly supported
- Independent source available
- Technical implementation details available
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?
Community
Comments
No published comments yet.