
Healthcare struggles because time is fragmented across too many systems that do not talk to each other. Point blank. At Welby, we stopped trying to optimize around that reality. We built something new instead. Markus OS.
Markus OS is our internal AI operating system designed to give our teams leverage. Not suggestions. Not summaries. Actual execution.
Built around our Markus MCP Server, a structured collection of custom Model Context Protocols that define what Markus can see, reason about, and do across our environment.
These MCPs are deliberate and opinionated. They give Markus real capabilities while keeping humans firmly in control.
We group them into three core areas:
Together, they form a real execution layer, not an abstraction.
A key unlock for us has been Markus’ native connectivity with n8n.
We design automations visually in n8n, then expose them to Markus as callable actions. Markus can trigger workflows, pass structured inputs, evaluate results, and chain multiple automations together based on context.
This shifts automation from static workflows to adaptive systems.
Markus does not tell someone what should happen next. Markus makes it happen.
All of this is surfaced through Markus Chat, the interface our internal teams use to work with Markus directly.
Nurses, ops, product, sales and leadership interact with the same system, each with the right context and permissions. No jumping between tools. No prompt gymnastics. Just intent, execution, and visibility.
Healthcare still lives inside browser-based systems, portals, and third-party tools that were never designed for automation.
So we built Markus Bridge, a Chrome extension that allows Markus to interact with the web the same way a human does. It can fill forms, click buttons, scroll, search, and move across tabs and windows.
This removes the gap between systems with APIs and the ones that still run the world.
Markus OS is not about replacing clinicians. It is about protecting their time and focus.
By collapsing reasoning, automation, and execution into a single operating layer, we give our teams leverage. Fewer handoffs. Less cognitive overhead. More consistency. More time spent on actual care.
We are not layering AI on top of healthcare. We are rebuilding the operating system underneath it.
And we are just getting started.