A patient who knows they need an orthopaedic surgeon can search a specialist directory. A patient who only knows that their symptoms are worrying has a different problem: they may not know which specialty to select. That uncertainty is where conversational navigation becomes commercially and clinically important.
On 13 April 2026, Doctify announced an AI Chat Assistant intended to help patients navigate towards an appropriate clinician. The company describes follow-up questions, specialty suggestions and urgent-attention signposting, while stating that the assistant does not diagnose or recommend treatment. These are intended functions, not independent validation results.
For doctors, the interesting question is how a conversational interface changes the route by which patients reach a clinic. That deserves more scrutiny than treating the launch as simply another website feature.
Navigation, triage and diagnosis are different jobs
A navigation tool may explain the difference between clinical specialties. A triage system assesses urgency or directs a person to a level of care. A diagnostic system considers possible explanations for symptoms. Those functions can overlap in a conversation, even when a product is marketed primarily as navigation.
A useful evaluation therefore asks what the assistant is allowed to do, what it declines and what happens when the user's problem does not fit a routine booking pathway. A disclaimer cannot, by itself, tell a clinician how well those boundaries work in practice.
Doctify's launch announcement should be read as a description of the company's intended service, not as independent evidence of sensitivity, safety or comparative clinical performance. A responsible article or practice briefing should avoid inventing validation results that are not supplied.
Why conversational discovery may affect private practice
A traditional directory gives the patient filters. A conversational assistant may help formulate the problem before those filters are applied. This could affect which specialties enter the patient's shortlist, how they describe their concern and which services they consider relevant.
That is an inference about the interface, not a claim about the assistant's undisclosed ranking rules. A doctor should not assume that buying a subscription improves clinical recommendations, or that a particular number of reviews guarantees visibility in an AI-generated shortlist.
The practical preparation is less exotic: maintain accurate scope, clinic locations, age limits, funding routes and appointment information. A profile that clearly distinguishes a consultation from a complete investigation pathway is useful whether a person or a software system reads it.
More enquiries may not mean better enquiries
Suppose a hypothetical clinic receives 30 enquiries through a new discovery route. Twenty are suitable, 12 book and ten attend. Another route generates 15 enquiries, 14 suitable, 12 bookings and 11 attendances. The smaller route may create less administrative work and more useful activity.
This example is not a Doctify result. It illustrates why a clinician should measure suitability and attendance rather than celebrating a larger inbox. The assistant could be commercially helpful if it reduces confusion, but disappointing if it simply routes more ambiguous problems into private booking requests.
Ask the practice team to record why enquiries cannot proceed: wrong age group, inappropriate specialty, unavailable funding, need for another pathway or lack of required records. Review aggregate patterns without creating an unnecessary database of sensitive conversational transcripts.
What clinicians should ask about safety
A serious product discussion should address urgent or deteriorating symptoms, ambiguous descriptions, language limitations and people who repeatedly seek reassurance. Ask how the system communicates uncertainty and how it directs people away from inappropriate booking when necessary.
Also ask about responsibility for the handover. Does the practice receive the original concern, a generated summary or only a booking request? Can staff distinguish what the patient wrote from what software inferred? Is there an accessible route for reporting unsafe navigation or inaccurate profile matching?
These are procurement questions rather than findings of a defect. They help prevent a common mistake: assuming that because a platform is familiar as a directory, a new AI-mediated clinical interaction needs no additional assessment.
Watch for commercial and clinical ambiguity
A patient-facing platform may have commercial relationships with participating providers. That does not prove that its navigation is biased, but it makes transparency about selection and presentation important. Clinicians should ask how paid visibility, available appointments and clinical suitability are distinguished.
The GMC's conflicts-of-interest guidance is relevant when a doctor's commercial interests could influence professional decisions. A practice should not allow a profitable acquisition channel to determine whether a consultation or intervention is clinically justified.
Consider a simple test: would the practice give the same advice about the appropriate next step if the patient were not a prospective paying customer? Commercially successful navigation should make appropriate care easier to find, not manufacture a need for a private appointment.
Protect information at the handover
Do not assume that consent to use a discovery chatbot authorises every downstream use of the conversation. Clarify what information is collected, where it goes, what the practice receives and which records the clinician must verify directly.
A booking summary should not silently become the authoritative medical history. At the consultation, confirm the presenting concern and relevant details through the normal clinical process. If a generated summary is retained, its provenance should remain clear to the people relying on it.
The same distinction applies to marketing analytics. Understanding which channel produced an appointment does not require placing symptom narratives into a general advertising system. The ICO's direct-marketing guidance is a relevant starting point when designing communications and data flows.
What this means for a doctor's toolkit
The likely strategic issue is not that AI will simply choose a doctor on the patient's behalf. It is that software may increasingly shape the questions patients ask before they choose. Practices should observe that change without treating speculative ranking theories as established facts.
Run a small, documented audit of enquiries from AI-mediated routes. Look for misunderstandings that can be corrected in the service description. Escalate safety concerns through the platform and local governance processes, and avoid drawing conclusions from a handful of self-reported experiences.
Clinician-facing reference tools address a different task. Ask iatroX can be used to explore clinical questions and inspect supporting material, with professional verification. It should not be presented as an equivalent patient-navigation product or evidence that another company's assistant has been clinically validated.
Related reading: AI search and specialist discovery and interpreting online doctor ratings.
FAQs
Is Doctify's AI assistant an AI doctor?
The launch describes assistance with finding an appropriate clinician. That should not be equated with independently validated diagnosis, emergency triage or autonomous medical care. Assess the actual intended function and published evidence.
Will paying for Doctify make an AI recommend me?
No such guarantee is established by the material reviewed. Ask how paid services and patient-facing recommendations are separated, and avoid assuming knowledge of undisclosed ranking mechanisms.
What should a private practice measure after the launch?
Track suitable enquiries, completed appointments, reasons for mismatch and additional administrative work. Report possible safety issues separately. Raw traffic alone cannot show whether navigation has improved the patient journey.
Explore clinician-facing clinical questions and sources with Ask iatroX.