Dr Harry Lykostratis, Founder, Lead Engineer and Chief Executive Officer of Open Medical, explains how a new approach using an ‘AI Orchestrator’ can address the challenge of unifying AI ecosystems within healthcare, creating a framework to unlock the power of AI today and prepare for the next generation of ‘agentic AI’ that can transform clinical and operational workflows.
Across the GCC, ambitious national health strategies rightfully identify the immense potential of AI software to support clinical decision-making, enhance productivity, and improve the patient journey. This optimism is well-founded, as AI tools are advancing at an incredible pace. However, the global health sector is facing an ‘applicability crisis’, struggling to deploy these tools at scale.
While technologies like ambient voice for clinical documentation show promise, they are largely an evolution of existing processes, not the revolution needed to unlock widespread transformation.
The challenge of scalability and applicability
What is holding healthcare back from leveraging more advanced AI? Clinical and digital health leaders face common hurdles: data governance concerns often limit AI models to small datasets, while complex integration with diverse Electronic Medical Records (EMRs) presents a significant technical and financial barrier. This makes it challenging to embed solutions reliably into organizational workflows.
We must resolve these foundational issues to prepare for the near future: the shift from individual AIs to agentic AI ecosystems.
Today, we use AI for single tasks. In the future, an AI ‘agent’ will orchestrate multiple tools and data sources to achieve a complex goal with minimal human input. In many industries, these agents may operate as ‘black boxes’, but this is not an option in healthcare. Every step must be transparent, auditable, and explainable to ensure clinical governance and patient safety.
The solution: An AI Orchestrator
Open Medical is addressing these challenges through our Ensiab Orchestrator. Our Ensiab platform is already built for workflow optimization and interoperability with proven experience integrating with major EMRs and clinical systems globally. Our additional AI Orchestrator capabilities will handle the data cleansing and pseudonymisation required to make clinical data securely available for AI models, simplify their integration into systems and orchestrate their use within workflows for maximum efficiency.
Our vision is a single, unified integration point. Any AI tool connected to Ensiab can be securely and rapidly integrated with hospital systems, allowing healthcare organisations to experiment with new models and quickly scale those that prove effective. To prepare for the agentic era, our orchestrator is built on four key principles:
- Parallel Processing: Simultaneously analyse diverse data types like EMR data, PDFs, and images using multiple specialized AI tools.
- Sequential Processing: Chain AI models, where the output of one becomes the input for the next, creating sophisticated automation.
- Routing: Direct outputs in the correct sequence to execute specific tasks within a defined clinical workflow.
- Feedback: Create a learning loop from outcomes, enabling continuous model improvement over time.
A practical application: Optimizing referral management
Consider referral management, a common bottleneck. An AI agent can triage patients far more effectively than traditional linear processes. Using our four-step framework, the Orchestrator could simultaneously process a referral letter, the patient’s medical history, and a medical image using different specialized AI models. The combined outputs are then fed into a final model to determine the most appropriate care pathway—be it a specialist, a physiotherapist, or a return to primary care. The decision is automated, and a summary is written back to the patient’s record, improving efficiency and patient flow.
Our vision is a single, unified integration point. Any AI tool connected to Ensiab can be securely and rapidly integrated with hospital systems, allowing healthcare organisations to experiment with new models and quickly scale those that prove effective.
Building a future-ready healthcare system
Open Medical does not develop AI models. We are model-agnostic, as we believe organizations should have the freedom to choose the best tools for their needs. Our focus is on creating the agile platform that allows these models to work together seamlessly.
Our ambition is to provide an orchestration layer that integrates with existing systems, taps into data securely, and routes it to various AI models, presenting the outputs meaningfully within the clinical workflow. This will empower healthcare providers to move from isolated pockets of AI to a system where AI is a core driver of efficient, timely, and high-quality care. By addressing the applicability crisis now, we can harness these revolutionary technologies to help deliver a world-class standard of healthcare across the region.
- For more information, visit: www.openmedical.co.uk
