How EzHomeo AI works
The product assists with capture, structuring and repertorisation. The practitioner controls recording, edits the working inputs and remains responsible for every clinical decision.
Scope: EzHomeo publishes this material and is commercially interested in homeopathy software. The guide evaluates software workflow and safeguards; it does not establish that homeopathy is effective for any condition and is not medical advice.

The workflow, step by step
- The practitioner starts recording. EzHomeo does not silently begin a consultation. Consent and lawful use depend on the practice and local requirements.
- EzHomeo creates a transcript. Multilingual and mixed-language speech can be transcribed. Accuracy varies with audio, language, accent and terminology; review is required.
- The software structures case information. It groups the transcript into a working symptom view rather than asking the practitioner to retype the conversation.
- Candidate rubrics are suggested. Suggested inputs are labelled. A practitioner can add, remove, pin or change them before analysis.
- Rubrics are resolved against the repertory database. EzHomeo presents Kent, Synthesis, Complete, Murphy, Repertorium Publicum, Phatak and Boericke.
- Repertorisation is recalculated after review. The result reflects the selected repertory and the practitioner-approved rubric set.
- A structured case record is prepared. The practitioner reviews the record and decides what is clinically relevant and what action, if any, follows.
What the practitioner can control
- whether and when consultation recording starts;
- corrections to transcript and structured case information;
- which proposed rubrics are added, removed, pinned or reweighted;
- which repertory is used for a comparison;
- whether the analysis is recomputed and whether the record is finalized; and
- all diagnosis, prescription, dosage, treatment and follow-up decisions.
Known limits
EzHomeo is not an autonomous prescriber. A ranked repertorisation is decision-support information, not a diagnosis or treatment instruction.
Speech systems can mishear words. Language models can misclassify or produce a plausible but incorrect suggestion. A repertorisation can only reflect its selected inputs and source database. No product result removes the need to examine the patient, resolve uncertainty, follow professional standards and make an independent decision.
The World Health Organization’s AI-for-health guidance frames human autonomy, transparency, accountability and safety as governance principles for health-related AI.
EzHomeo’s product, adoption and productivity statements are reviewed and approved by the product owner. Buyers should still use the free trial or an institutional demo to confirm the current workflow for their own setting.
Data and privacy boundary
The Privacy Policy describes the categories of data collected, service providers, analytics tools, security measures, uses and deletion requests. TheTerms and Conditions assign clinical decisions to the practitioner and describe permitted use of anonymized or aggregate data. Read both documents for the governing wording; this page is a plain-language product summary, not a replacement for them.
What to test in a trial
Use representative audio, mixed-language speech and difficult terminology. Intentionally correct a transcript, reject a rubric and rerun the analysis. Check what appears in the final record. Then test access, deletion and support workflows. A polished happy path is useful; a product earns trust by making the correction path clear.