In today’s digital economy, data is the new currency—and businesses that know how to extract, refine, and leverage it effectively are the ones gaining a competitive edge. In the healthcare sector, the stakes are even higher. The ability to transform data into actionable insights and seamless processes not only drives efficiency but directly impacts patient outcomes and member satisfaction. For medical scheme administrators and managed care organisations, this means rethinking how they handle interactions, especially when beneficiaries reach out in times of medical need.
Healthcare scheme beneficiaries often only engage with their medical schemes during moments of vulnerability—when seeking preauthorisation for hospitalisation, querying benefits, or managing diagnoses. These interactions are time-sensitive, emotionally charged, and often complex. At Medscheme, we recognised that enhancing these experiences would require more than empathy—it would require automation, driven by clean, structured data.
Laying the Foundation for Hyper-Automation
Medscheme’s journey to automation began over 7 years ago with the introduction of Robotic Process Automation (RPA), designed to offload repetitive, manual tasks and free up skilled teams to provide direct support to members. Today, over 65 business processes are automated, with 100 bots now managing nearly 200,000 work items per month.
However, scaling automation required us to confront a central challenge: unstructured data. Emails and free-text web portal requests from hospitals—rife with medical shorthand and varied formats—dominated our servicing channels. To drive true efficiency, we needed to turn this unstructured data into structured inputs that bots and systems could process accurately and quickly.
Solving the “Data In” Challenge
Our “data in” strategy became the core enabler of high-level automation. With the understanding that changing customer behaviour overnight was unrealistic, we opted for a dual approach—enable structured self-service digital channels where possible, but also build solutions that can ingest and interpret unstructured data in traditional servicing channels like email and web portals.
In collaboration with AfroTech and Cogent, our Hyper-Automation team developed an intelligent solution stack that included:
- Robotic Process Automation (RPA): Automating web portal logins, interpreting requests, and completing transactions, all while reducing human error.
- Natural Language Processing (NLP) with Azure Cognitive Intelligence: Automatically detect text of a given language, understand the sentiment and structuring data form unstructured data inputs.
- Agentic AI technology (Stubber):
Transforming shorthand, social media-style text, and clinical language into structured data by extracting key details such as ICD-10 codes and service dates.
- Microsoft Logic Apps: Orchestrating data flows, choosing appropriate extraction models, and ensuring seamless coordination across platforms.
- Secure Web Services: Ensuring compliance and privacy by encrypting data transmission.
- Data Validation Rules: Reinforcing accuracy by applying strict business logic before transactions are completed.
Tackling Complexity with Agility
Bringing this ecosystem to life in a healthcare environment presented unique challenges—from decoding informal shorthand to integrating multiple technologies under stringent data regulations. Through an agile development approach, these complexities were broken down into manageable iterations. Continuous testing, stakeholder feedback, and real-time refinement enabled the team to build robust, scalable solutions.
Real-World Use Cases
Our automation efforts have focused on three high-impact areas:
- Preauthorisation Requests and Updates (Renal, Radiology, Physiotherapy, Hospitalisation, Mental Health): These now experience faster processing with fewer errors.
- ICD-10 and Tariff Code Assignments: Automation ensures consistent clinical and billing data assignments, enhancing compliance and accuracy.
- Data Validation and Transaction Completion: Ensures end-to-end data accuracy and seamless transaction finalisation.
Tangible Outcomes
The impact has been profound:
- 90%+ data accuracy from unstructured inputs
- 85% automation rate, delivering 20,000+ straight-through processed transactions per month
- Improved operational efficiency, enabling skilled resources to focus on member support
- Reduced servicing times and human error, enhancing member experiences
- Consistent, compliant code and tariff assignment
- Reliable claim completion, ensuring transactional integrity
Driving the Future of Member Engagement
Through this transformative journey, Medscheme has redefined what efficient, member-centric healthcare administration looks like. By harnessing data, intelligent automation, and advanced technologies, we’re not just improving internal efficiencies—we’re actively reshaping the service experience for customers of our client Medical Schemes.
As the healthcare landscape continues to evolve, so too will the technologies that support it. Medscheme remains committed to staying ahead of the curve, using data and innovation to ensure that every member interaction is not just a transaction, but a step toward better, more coordinated care.
