Sari la conținut

Klu for clinics

Acest conținut nu este încă disponibil în limba selectată.

Klu is the assistant built into every KlusAI product; inside Med it is presented as Med, the healthcare assistant. It answers from your own clinic’s records and documents rather than from general knowledge, and cites what it used.

This page is reference material. If you have not used the assistant before, the quickest introduction is to open it, type /, and pick a command from the list.

Type / in the message box and choose from the list, or keep typing to filter it.

Command Shown as What it does
/appointments AppointmentsShow today’s scheduled appointments Lists today’s appointments chronologically as Time — Patient — Type — Status, and highlights anything due in the next two hours.
/patient_summary Patient summaryComprehensive patient overview Finds the patient, then presents demographics, recent visits, key notes from the record, and upcoming appointments.
/clinical_guidelines Clinical guidelinesSearch clinical guidelines and protocols Searches the documents in your clinic’s knowledge library, cites the specific sections and recommendations it used, and notes the evidence level and strength of recommendation where the source states them.

/appointments and /patient_summary need assistant access; /clinical_guidelines needs your clinic’s knowledge library to be enabled. Commands you cannot use are not shown.

On an empty conversation Med offers four starting points, which are just messages you could type yourself:

  • Search the medical knowledge base
  • Today’s appointments
  • Search for a patient
  • Summarize clinical guidelines

Within your clinic only:

Capability What it means in practice
Search patients Finds patients by name, email or medical record number.
Search appointments Finds appointments by title, patient name or status.
Search the knowledge library Retrieves passages from the clinical documents your clinic uploaded, with citations.
Cross-reference documents Compares documents you point it at against each other.
Extract structured data Pulls named fields out of a document into a structured answer.

Everything is scoped to the clinic you are working in and to what your role lets you see. The assistant has no access to another clinic’s patients, and no access to the public internet or to a general medical corpus — if the answer is not in your clinic’s records and documents, it does not have it.

  • It will not diagnose, stage, or recommend a treatment.
  • It will not generate or sign consent documents. That is the consent engine, and it runs through a clinician’s review — see How the consent engine works.
  • It will not invent a source. If it cannot find support in your documents, it says so.
  • Name the patient precisely. Where several patients share a name, add the date of birth; the assistant searches text, and it cannot tell two same-named patients apart for you.
  • Upload the guideline you want cited. /clinical_guidelines searches your clinic’s own library. Add protocols under Clinical knowledge → Documents; they are indexed automatically.
  • Check the citation, not the summary. Every knowledge answer points at the passage it came from. Open it — that is the point of the citation.
  • Ask one thing at a time. A patient summary followed by a guideline question works better than both at once.
  • Approved consent content — the same knowledge library also powers passage suggestions when you author consent wording.
  • What is kOS? — how knowledge bases, organizations and roles work across KlusAI products.