Responsibility Meets Innovation: Hamilton Medical’s Path to Scalable AI

Hamilton Medical faced rising process volumes, strict regulatory requirements, and manual processing steps that limited scalability.

As a manufacturer of medical devices, every step - especially those involving AI-driven procedures - had to be compliant, traceable, and fully auditable across ISO, FDA, and EMA standards. Meanwhile, manual workload continued to increase. Orders were manually transferred into the ERP. Translations remained in the pipeline for weeks. And compliance teams evaluated incident reports individually.

Early discussions made it clear this would not remain a single project. Different departments contributed new ideas - from document classification to automating regulatory tasks. The result led to numerous parallel initiatives combining technical, organizational, and regulatory aspects.

In collaboration with SoftwareOne, the Medtech company decided to restructure its approach to AI and automation altogether. The objective became building both technical and organizational foundations for long-term success.

From the initial projects, a holistic AI ecosystem emerged. The goal was not only to automate individual processes but also to establish a scalable and transparent structure for sustainable AI adoption. The projects were designed to complement one another, forming a stable foundation for future innovation.

1. Automated Order Processing – A Global Core Process Becomes Scalable
   The first milestone was automating order processing, a globally critical process. Incoming documents in various formats are now read by an LLM-based AI service that extracts relevant data, matches it with master data, identifies potential errors, and transfers the information directly into the ERP system.
   
   In six weeks, the team developed an initial solution that increased efficiency by approximately 50%. With further optimizations, including prompt design, model configuration, and end-to-end integration, efficiency increased to approximately 70%.
   
   “An essential success factor was our structured approach to technical decisions. Switching from Azure Intelligence Documentation to flexible LLM-based pipelines was a major learning curve. We consistently made joint decisions and evaluated new requirements methodically.”
   — Daniel Ramseier, Head of AI and Business Automation, Hamilton Medical AG
   
   The team implemented a multi-vendor strategy in parallel. In addition to Microsoft, they evaluated Google and AWS options to reduce technological dependencies and ensure a scalable architecture. The open-platform architecture supports this flexibility.
   
   Another key contributor to project success was the involvement of the partner Concentrate. Their solution enables stable connections between the ERP system and the AI layer, accelerating implementation. Standardized integration mechanisms reduced customization effort, shortened testing cycles, and improved overall system stability.
2. AI-supported Translations – A Critical Regulatory Step Accelerated
   Translating medical documents had long been a time-consuming and costly process for Hamilton Medical. A new AI-based workflow is being implemented, built on Azure technologies, capable of processing XML content, preserving formatting, and leveraging previously translated content. The system assesses translation quality and can also be used for marketing and eLearning content.
   
   Crucially, the process continues to comply with ISO 17100 standards. AI does not replace regulatory review but will significantly shorten throughput times and ease the burden on departments once fully deployed.
3. Sanity Check: Automated Compliance Review
   Hamilton Medical is now working with SoftwareOne on an AI-driven solution to process device incident reports. The system will pre-classify incoming reports, extract essential information, highlight potential risks, and support the creation of regulatory documentation.
   
   Final decision-making remains with the compliance team, but automated pre-processing significantly reduces their workload. The project is currently in the implementation phase, focused on refining classification and integrating additional regulatory requirements.
4. Shared Infrastructure
   Alongside the functional projects, a scalable technical and organizational foundation for AI usage is being established. These include cloud architecture, DevOps pipelines, security concepts, permissions, logging, and monitoring. The individual projects interlock and contribute to building a long-term environment where Hamilton Medical can use AI responsibly and at scale.

What Has Changed Through the Use of AI

The introduction of the AI ecosystem changed not only how processes run but also how innovation is structured and implemented within Hamilton Medical.

1. Significant Relief in Daily Operations
   AI-driven order processing reduced manual processing time per order by approximately 80%. Routine tasks that once required multiple employees now run automatically in the background, freeing resources for complex work.
2. Higher Process Quality and Consistency
   AI-supported workflows now deliver more structured data, consistent documents, and improved regulatory traceability. Error-prone manual steps have decreased, increasing overall process quality.
3. Shorter Throughput Times in Key Business Areas
   Orders are processed faster. Translation workflows are more predictable. Compliance teams can focus on complex cases instead of sorting and preparation tasks.
4. A Scalable Base for Future AI Projects
   Hamilton Medical now has modern cloud infrastructure, governance mechanisms, and an internal team capable of maintaining and developing AI applications. New projects can be initiated and integrated quickly.
5. A Portfolio That Continues to Grow
   Over 25 AI initiatives are now running in parallel or in preparation. Innovation is no longer the exception but a continuous, structured process.