Building and evolving the business-critical microservices behind Clinic24 — the SaaS platform combining AI-driven patient triage with digital consultation infrastructure for Sweden's primary healthcare market.
Sweden's primary healthcare system operates under persistent pressure. Demand consistently outpaces available appointment capacity, and the triage process — deciding which patients need what level of care, how urgently — is simultaneously one of the most consequential decisions in primary care and one of the most time-intensive to perform manually. Clinics and regional healthcare partners needed a way to extend capacity without compromising the quality of clinical decision-making at the front door.
Doktor24's answer was Clinic24: a fully integrated SaaS platform designed specifically for the Swedish primary care market. At its core, a capability that no competitor had delivered as a fully functional production system — an AI-powered chatbot that performs intelligent patient triage, assesses the nature and urgency of a patient's need, and routes them to the appropriate level of care. Wrapped around that: a complete digital healthcare environment including video consultation infrastructure for real-time patient interaction with doctors and nurses, medical data management, and a platform architecture flexible enough to serve regional healthcare partners with varying operational requirements.
The engineering challenge was not a one-time build with a clear finish line. Clinic24 was a living platform — continuously receiving new requirements from practitioners and regional partners, operating under the reliability expectations of a clinical system, and serving a user base that included doctors, nurses, engineers, and patients simultaneously.
Doktor24 is unique in the Swedish market in operating a fully functional AI chatbot for patient triage as a production system — not a prototype or a pilot. The chatbot handles the initial clinical routing decision for patients entering the platform, directing them to the right care pathway before any human clinical resource is engaged. Engineering the infrastructure that supports that capability, reliably and at scale, is the foundation everything else depends on.
We joined the Clinic24 engineering team as Senior Software Engineer, taking ownership of designing and implementing the business-critical microservices at the centre of the platform — the services managing patient and medical data, and the infrastructure enabling digital consultations between patients and healthcare providers.
The patient data microservices sat at the intersection of clinical process and software architecture. They handled medical records, appointment flows, triage outcomes, and the data exchanges that connect patients to the right care pathway. Working at this layer in a clinical system means there is no tolerance for data integrity failures — an error here is not a bug report, it is a patient care event. The services were designed accordingly: robust, well-validated, and built to handle the asynchronous complexity of real-world patient journeys without losing consistency.
The video consultation infrastructure was the other primary workstream. Enabling real-time interaction between patients and healthcare providers — reliably, across a distributed user base — required careful WebSocket management, session handling under variable network conditions, and the kind of resilience testing that surfaces edge cases before patients encounter them in a clinical context.
The deployment environment was Kubernetes on Google Cloud, and we contributed directly to the configuration, load balancing, and performance testing that kept a multi-tenant SaaS platform stable under production load. The team operated under a GitOps model — deployments version-controlled, reproducible, and auditable — appropriate for a system serving regional healthcare organisations with strict operational and compliance requirements.
Throughout, the work was genuine continuous delivery: translating an ongoing stream of requirements from practitioners, engineers, and regional partners into production software, while maintaining the technical quality standards that clinical infrastructure demands.
The engagement delivered continuous improvement to a platform that regional healthcare partners and patients depended on daily — not a single release, but a sustained uplift to the reliability, capability, and scalability of a production clinical system.
Tobias was always very professional in his approach, and offered a multi-faceted take on many of the complex ideas and solutions that he was tasked with. Working in the Clinic team required a great deal of creative thinking and on-the-fly decision making — in this regard Tobias was a real asset, offering good discussions and ideas based on his wealth of knowledge and experience.
The bottleneck in primary healthcare is rarely the quality of care once a patient reaches a clinician. It is everything that happens before that — the time spent deciding who needs to be seen, how urgently, by whom. Done manually at scale, that process is slow, variable, and resource-intensive. Done well with AI, it routes patients faster, frees clinical time for decisions that genuinely require clinical judgment, and reduces the risk of under-triage for patients who need immediate attention.
Doktor24 built that capability as a real, production system — not a pilot, not a chatbot that escalates everything to a human after three questions. Engineering the platform infrastructure that makes that possible — the data services, the consultation layer, the deployment environment — is exactly the kind of work where technical quality has a direct clinical consequence. A platform that degrades under load is not a performance problem. It is a patient access problem.
That is the standard we brought to the Clinic24 engagement: the same engineering discipline we apply to any system where the users are real, the stakes are real, and the difference between a well-built system and a poorly-built one is measured in something more important than uptime metrics.
If you're building a digital health product where reliability and patient outcomes are inseparable — we'd like to hear about it.
Start a conversation