Our Thesis
Sanctuary Thesis
We want to bring about healthcare abundance through the acceleration of AI in healthcare. We believe AI is the vehicle through which we can rejuvenate a creaking health system. This document explains why and how.
The Problem with Software in Healthcare
Healthcare is intrinsically messy, heterogeneous, and deeply context-dependent. This has historically made it a poor fit for software, which rewards standardisation and uniform scale. Effective digital systems in healthcare often need to be semi-custom or locally adapted - a model that was previously too slow, expensive, and fragile to deploy widely. As a result, technology either failed outright or collapsed into monolithic platforms that entrenched complexity rather than reducing it.
How AI Changes This
We see AI changing this dynamic in two fundamental ways:
The cost of writing and deploying code is collapsing.
This has important downstream implications. What was once economically prohibitive—deploying and iterating tailored systems across many local contexts—is now within reach. This does not eliminate complexity, but it makes it workable. It allows technology to conform to clinical reality, rather than forcing clinical work to conform to technology.
AI agents can increase healthcare supply.
Healthcare is and always has been supply-constrained, resulting in delayed appointments, diagnoses, and treatments. Agents present the first real ability to connect historically disparate systems in a way that only humans could do previously. Within the near future, it should be possible for the entire system to be relieved of the core administrative burdens that currently consume it, freeing that resource to be reallocated to actually delivering care.
In this sense, healthcare may experience a form of Jevons paradox: as efficiency improves in non-clinical domains, demand does not simply rise in volume, but in quality. Freed capacity can be reinvested in longer consultations, better continuity, and more humane care—outcomes that scarcity has progressively eroded.
Our Approach
We will achieve this by deploying deeply technical talent into clinical settings, understanding the problems faced on the ground, and rapidly customising our core AI infrastructure to fit these contexts.
Our first focus is on new modalities of data collection. AI transcription, voice AI, and chatbots present entirely new ways for clinicians and patients to interface with healthcare—and for startups to harness and collect data. Sanctuary's starting point will be deploying AI solutions into administrative bottlenecks: phone calls, referrals, triage, and more.
For too long, healthcare professionals have carried the burden of a failing system through personal sacrifice, improvisation, and exhaustion. Patients, in turn, have been asked to accept delay, fragmentation, and indignity as unavoidable. Neither is acceptable, and neither is inevitable.
If we can fix healthcare, then we must. Not because it represents a compelling business opportunity, but because the cost of inaction is measured in harm that could have been avoided.
The moment calls not for another review or reform cycle, but for the willingness to build the future we want.
Let's talk
Whether you're a clinic drowning in admin, an NHS trust looking for AI partners, or a health-tech company exploring what's next — we'd like to hear from you.
Get in Touch