Mitsubishi Tanabe Pharma streamlines cloud security operations and employee literacy
Mitsubishi Tanabe Pharma, a leader in the pharmaceutical industry, sought to improve its cloud security operations amid a company-wide push for digital transformation and data utilization. The company was challenged by the increasing complexity of cloud security, especially as workflows expanded into precision medicine and patient-centric solutions. Security incident checks and alerts primarily in English made rapid response and understanding difficult for its Japanese-speaking team. To address these issues, the firm undertook a comprehensive assessment of its Azure-based infrastructure with Microsoft guidance, adopting Microsoft Defender for Cloud and Microsoft Sentinel for consistent policy enforcement and continuous monitoring. The real breakthrough came with the integration of Azure OpenAI Service, which automated the translation and summarization of security alerts into Japanese and provided clear explanations of threats and recommended actions. This reduced incident handling time tenfold and significantly bolstered employee security literacy. Now, security incidents are handled more efficiently by a lean team, supporting the company's 'VISION 30' strategy for advanced, secure digital healthcare initiatives.
- Organization
- Mitsubishi Tanabe Pharma
- Industry
- Pharma
- Location
- Japan
Reported outcomes
550 hours
timeTime & speed
Strategic outcomes
Primary read
Use case focus
Showing 3 of 3
- 1Automated Security Incident Translation and Summarization for Cloud Operations
- 2Real-Time Alert Explanation for Cloud Security
- 3Integrated Cloud Security Monitoring with AI-Driven Language Processing
- Incident check times averaged 10 minutes each, delaying rapid response.
- Security recommendations and alerts were only available in English, impeding understanding for Japanese staff.
- Growing complexity and volume of cloud-based operations increased the risk of misconfigurations or missed threats.
- Limited human resources required that security processes be highly efficient.
- Data utilization for precision medicine and patient care needed a robust, secure cloud foundation.
- Implemented Microsoft Defender for Cloud and Microsoft Sentinel for automated detection, assessment, and continuous monitoring.
- Adopted Azure OpenAI Service to translate and summarize English security notifications into Japanese, explaining the significance and risk of each alert.
- Integrated security operations, enhancing visibility and policy enforcement based on Azure Well-Architected Framework recommendations.
- Automated delivery of digestible, incident-specific security summaries to staff via email and Microsoft Teams, reducing manual review.
- Received technical guidance and knowledge transfer from Microsoft for successful cloud security and AI integration.
- Reduced incident check time from 10 minutes to 1 minute per case, saving approximately 550 hours annually.
- Significantly improved security literacy and risk awareness among employees.
- Enabled efficient cloud security management with minimal human resources.
- Enhanced ability to respond quickly and effectively to security incidents, directly supporting digital transformation objectives.
- Strengthened overall security posture of the company's Azure-based infrastructure.
Architecture
The solution architecture includes Microsoft Defender for Cloud as a central platform for managing security policies and workloads, integrated with Microsoft Sentinel for SIEM functions—collecting and analyzing security logs, managing identity-based access, and issuing automatic alerts. Azure OpenAI Service is used to process incident messages: translating, summarizing, and annotating security alerts into Japanese, then distributing them via email and Microsoft Teams. The automation leverages Azure Logic Apps for workflow orchestration, connecting Defender for Cloud and Sentinel incident data to OpenAI and onward notification channels.
Sources & evidence1
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?