Reinventing CSL through cloud transformation and agentic AI

CSL faced significant complexity from accumulated technical debt in their data centers. Through multiple acquisitions, it was faced with operating disparate technology stacks with a mix of licenses and vendors that ultimately presented recurring capacity and scalability issues. One of the core concerns was modernizing 5000 VMware servers and over 1000 applications across 29 data centers.

“We took a step back and decided how we wanted to reinvent CSL. Our goal was to create a robust environment that fosters innovation while keeping things simple, cost-effective, and productive,” said Mark Hill, Chief Digital Information Officer at CSL. “We set out to build a technology foundation that’s agile and resilient—one that makes it easier for our people to do their best work and make CSL an even better place to work focusing on efficiency. We created our Technology Modernization Program (TMP) to invest in the skills our people need for the future. We’re embracing emerging technologies, including AI, to drive innovation and stay competitive. We want no one left behind in this transition – we’re upskilling and supporting our people to confidently play their role.”   

The program also optimized how CSL works through streamlined workflows and automated processes. “We wanted to foster seamless collaboration, shared ownership, and to enable teams to operate as one,” said Erik Hong, Global Head, Enterprise Systems & Processes, CSL. “We wanted to make data more accessible and actionable to help transform our talent.”   


CSL aimed to revolutionize its technology infrastructure through strategic modernization and technical debt reduction. This transformation is needed to enable the company to deliver more efficient and reliable services while leveraging scientific innovation, cutting-edge technology, and human expertise to deliver lifesaving medicines and drive sustainable growth. “Our end goal is to create therapies that deliver enduring patient impact—innovations that make a lasting difference in people’s lives,” said Hong.

CSL has established its modern cloud platform to achieve its goals. Through this CSL would standardize access to cloud services, security, compliance by design, end-to-end service orchestration, and AI readiness. The company chose AWS to  lift and shift its estate delivering a secure, resilient, and data driven environment, transparent monitoring, and measurement across the system. It also reduced dependency and cost of on-premises data processing. This new solution on AWS would deliver enterprise data governance, align with organizational goals, and reduce complexity via fewer applications, scale, transparency, and reduced licensing cost.

The company planned to reduce its data center footprint but traditional wave planning and server migration methods proved inefficient. The traditional process required extensive manual effort due to scattered unstructured documentation and complex dependencies across data centers. Initial planning showed poor scalability - taking one month to plan 172 servers and one to two hours per application discovery, making it unsuitable to meet project timelines.

“As part of TMP, one of our strategic initiatives represents a partnership between CSL and AWS ProServ aiming to modernize systems, reduce technical debt, and deliver faster, more reliable, and secure services while empowering our workforce through upskilling and technology enablement. TMP delivers measurable business outcomes focusing on migration to Digital Core - our branded, secure landing zone environment for all things AWS, modernization, and retirement of legacy systems. Moving to AWS also advances our sustainability goals, as we’ve committed to a substantial reduction in absolute Scope 2 emissions by 2030,” said Hill. Scope 2 emissions result from activities that generate electricity consumed by the facility, such as heating and cooling. 

AWS is the foundation for CSL's Digital Core, which consists of Secure Cloud Foundation, Data and Analytics Platform, and platforms and services to accelerate migration of applications. CSL and AWS Professional Services used AWS Transform for VMware to automate wave planning and server mapping. AWS Transform for VMware uses specialized agents to analyze and map complex VMware environments, converting network configurations into AWS built-in constructs, and helps to orchestrate dependency-aware migration waves for seamless cutovers. With the agentic AI capabilities of AWS Transform, CSL coordinated a phased migration approach aligned with system readiness and business continuity. AWS Transform agents documented the relationships and dependencies between on-premises servers, applications, and data, and used cloud targets to ensure continuous uptime during migration. Transform then grouped applications and infrastructure into manageable phases or ‘waves’ for migration. 

“We fed 4.6 million records of on-premises discovery data residing on over 5000 VMWare servers across 29 data centers to AWS Transform, processed by the wave planning agent to generate an initial wave plan, saving months of manual effort.” 

Mark Hill
Chief Digital Information Officer at CSL

CSL is on track to exit two major data centers by December 2025. The Data and Analytics Platform will be operational with production environments deployed across three AWS Regions, supporting critical business initiatives including R&D, Donor Health Vault, and Deviation AI, which uses AI-powered analytics to find deviations or outliers in trial outcomes, resulting in faster evaluation and adjustments to trial protocols.

CSL used Amazon Q Business to index infrastructure requirements documents spanning over 1000 applications with more than 17,000 pages. The documents contain vast and deep knowledge about all the applications and on-premises servers, which the company fed into Amazon Q Business. AWS Transform for VMware generated the initial wave plan, and the three Amazon Business applications automatically created the initial application discovery reports in minutes compared to hours of manual work.

CSL launched Project Unify to consolidate its fragmented ERP landscape as part of its broader TMP. The company operated three separate SAP instances, creating inconsistent data standards and fragmented governance. Project Unify will establish ‘One CSL’ through AWS for RISE with SAP. RISE with SAP bundles SAP S/4HANA Cloud ERP, private edition, cloud infrastructure, managed services, and more in a single license. By December 2025, CSL will have completed two of three planned deployments, establishing enterprise-wide data standards and governance with globally common best practices. The project goal is the successful and seamless transfer of existing systems, prioritizing business-critical CSL subsidiary applications. “The move to AWS for RISE with SAP is consistent with our goal of avoiding managing infrastructure. Our AWS partnership has created a new enterprise governance structure and change leadership model, positioning us to operate as one unified team with industry-leading processes, data, and technology—ultimately improving the positive impact for patients and public health,” said Hong.

• Accelerated initial wave planning by 10x: Used AWS Transform VMware to save a minimum 10.5 weeks of effort for 1072 applications in 29 data centers.
• Ability to build and deliver new AI solutions: Leading the next era of innovation – powered by insights and tools that smooth and speed discovery and decisions. Working seamlessly with people and technology to deliver life-saving medicines and protect public health.
• Accelerated application discovery time by 12x for each application: Created three Amazon Q Business applications leveraging Infrastructure Requirements documents with Agentic AI wave plans to create systematic migration documentation with superior accuracy, uncovering critical data points that drive migration strategy. It brought down the application discovery process time to five minutes from one to two hours per application. 
• 30 months to exit from 17 data centers: Generative AI and Agentic AI approaches make CSL’s audacious goal a reality.

“With AWS, we built a robust foundation combining secure access, cloud infrastructure, and unified data management, designed for scale, transparency, and resilience. Modernizing legacy applications and data accelerates our ability to build AI solutions enables us to deliver life savings medicines to patients faster, ” said Hong. Over a 20x improvement in application discovery time and exiting on-premises datacenters in only 30 months revolutionizes CSL’s operations and enables it to upskill its talented workforce. On top of that CSL is able to realize about 30% operational cost savings through licensing cost avoidance, hardware costs, and datacenter related costs including floor space, energy, maintenance, and people.