A patient begins taking a new antidepressant in September. In much of England, the next appointment with the prescriber may be ten or twelve weeks away, sometimes longer. During that interval, the medication may begin to work, or it may not. Sleep patterns change, side effects emerge and mood can deteriorate. Yet the only person witnessing these developments may be the patient. That silence between appointments is the commercial opportunity Siggy, a New York start-up, hopes to address after raising $6 million.

The pre-seed funding round, led by Village Global, will support a service combining an AI intake conversation, a daily one-minute check-in and, shortly, readings from wearables such as an Apple Watch or Oura ring. The information is condensed into a score out of 100, which both patient and clinician can see each day. Siggy operates in the US through insurance and remains at an early stage. The gap it has identified, however, is familiar to mental health services on both sides of the Atlantic.

The model is straightforward. A free AI conversation gathers a patient's history, while diagnosis and prescribing decisions remain exclusively with a licensed clinician. Between appointments, the software asks patients about their sleep, mood and medication. If it identifies a pattern developing over several days, it alerts the clinician.

Siggy's founders bring experience from across healthcare and technology. Jonathan Sabbagh built Journey Clinical into a large network for psychedelic-assisted therapy after two decades in finance and discovering his own undiagnosed PTSD. Edvard Engesaeth co-founded Nurx, a prescription service. Several investors are backing the founders for a second time. Siggy accepts insurance across all 50 US states, including Medicare Advantage, and offers its service free to veterans. Wearable integration is expected shortly.

No NHS trust is buying Siggy's product. Its relevance to Britain lies in the problem it exposes. The NHS faces the same gaps between appointments, but has considerably less capacity to bridge them. Long waiting lists, persistent psychiatrist vacancies and stretched community teams leave patients with limited contact between consultations. The 10 Year Health Plan promises a shift from analogue to digital, with the NHS App intended to become the front door to care. Yet one crucial operational question remains unanswered, what happens when continuous monitoring produces information that somebody must act upon?

The promise of monitoring is that it can increase clinical capacity. That claim deserves scrutiny. A daily score for every patient on a caseload creates a continuous stream of signals, each potentially requiring a decision. In an insurance-funded market, clinicians are paid per encounter, meaning a flag can generate a billable appointment. Under NHS block contracts, the same warning creates additional work without additional revenue. If nobody has time to examine it, the record may show that the service knew a patient was deteriorating but failed to respond. That could leave a trust in a more difficult position than having no warning at all. Responsibility between the provider, integrated care board and software supplier also remains unsettled.

Data governance presents another significant obstacle. Daily reports covering mood, sleep and medication adherence, combined with biometric readings, constitute special category data under UK GDPR. Together, they would create one of the most intimate records held about any citizen. The collapse of care.data and the 2021 pause of the GP data for planning programme demonstrated how quickly public confidence can disappear when services move faster than their explanations. Any UK deployment would need to satisfy the Digital Technology Assessment Criteria and the clinical safety standards governing NHS software. It would also need to establish whether a proprietary score prompting clinical review falls within the MHRA's medical device rules. The answer may well be yes.

The evidence supporting the technology is less convincing than the funding round might suggest. A score out of 100 is a construct developed by the company itself. Siggy's launch material identifies a genuine problem but provides no peer-reviewed validation of the score, evidence a commissioner would expect before accepting it as a clinical instrument. NICE's early value assessment route allows promising digital products to generate evidence under supervision, making a tool of this kind a potential candidate. Investor enthusiasm for clinician-led AI in mental health, reflected in several comparable funding rounds this year, demonstrates commercial interest. It says little about patient outcomes.

For NHS leaders, the priority is to settle the practical questions before suppliers arrive. Mental health trusts and integrated care boards must establish which clinician is responsible for responding to a flag and what response time is realistic with existing staffing. Decisions about where the data is held follow from that. Policymakers also owe the market clearer guidance on when remote monitoring becomes a regulated medical device. Ambiguity favours large incumbents able to afford legal advice. British health-tech companies developing similar products face an American competitor backed by a reimbursement model that pays for engagement, while domestic buyers are unlikely to spend much before the evidence exists.

Patients could face another, quieter inequality. Those able to afford wearable devices will generate richer streams of information and potentially receive earlier warnings. People with severe mental illness, who may be most vulnerable to deteriorating without being noticed, are among those least likely to own such devices.

Siggy may succeed, although most early-stage companies do not. Its significance for Britain lies in the challenge it presents to a health service moving towards digital continuity faster than it is developing the capacity to respond. A daily score can tell clinicians that a patient is deteriorating. Acting on that warning still requires a clinician with time available. Before ministers commit the NHS to this technology, the supply of those clinicians is the figure they should be watching.