Homeo AI vs Rubrii: Local Repertory or Review Copilot?

Quick answer

Choose Rubrii when a practitioner wants a focused browser repertory workflow for finding rubrics, building a weighted case, viewing a repertorisation matrix, opening materia medica, and exporting a report. Choose Homeo AI when the work begins with a consultation narrative and requires an editable transformation into symptoms, proposed rubrics, inspected source passages, a practitioner-owned differential, and follow-up reasoning.

Rubrii's public guide presents a direct repertory sequence. Homeo AI presents a review copilot sequence. The practical difference is whether the practitioner already knows the rubrics to add or needs help preserving and reviewing the path from the patient's words to those rubrics.

Rubrii also makes a notable local-data claim. Its current homepage and guide say cases or patient data stay on the user's device. Its privacy policy, last updated in January 2025, separately says the service may collect search queries, repertorisation activity, case files, and patient information if provided. A clinic should not guess how those statements fit together. It should test the current app and obtain written clarification before entering identifiable data.

What the current Rubrii pages describe

Rubrii's official homepage positions the service as a modern repertorisation and materia medica platform for clinics and classrooms. The vendor currently advertises AI-powered search, a case panel, a repertorisation matrix, source badges, materia medica access, and PDF or shareable exports. It also displays large catalogue and availability figures. Those numbers are vendor claims and should not be treated as an independently audited source manifest.

The official user guide describes four steps: search for rubrics, add them to the case panel and adjust weights, compare remedies in a matrix, then export or share the result. This is a clear and testable workflow. A practitioner can take a known case, select a controlled rubric set, and compare Rubrii's matrix and export against a trusted reference.

The current terms, dated January 2025, describe Rubrii as an educational and professional reference platform for practitioners and students. They say users remain responsible for clinical decisions, information should be checked against authoritative sources, and emergencies require immediate professional attention. The terms also say free and premium plans may exist and that price changes will be communicated in advance. Because the live pricing route did not expose readable public plan details during this review, this article does not infer a current price.

Where Rubrii is genuinely strong

Rubrii's strength is a compact repertory workflow. When a practitioner already has a disciplined symptom totality, the case panel and weighted matrix can make repertorisation fast. Source badges and direct movement from rubric search to materia medica can reduce navigation. A clean export can also be useful for a teaching exercise or a clinic record after professional review.

The local-case design, if it operates as described, is another meaningful advantage. Keeping the working case on the device can reduce server-side exposure and make a repertory tool easier to evaluate with sensitive material. It may also appeal to practitioners who do not want every repertorisation stored in a cloud patient-management system.

Those advantages should be verified rather than dismissed or assumed. Test what is stored in browser storage, what is transmitted during search, whether an account is required, whether an export contains identifiers, what happens across devices, and whether clearing site data removes the working case. A local interface can still make network requests, and a cloud-hosted search can still receive query text.

The privacy pages need reconciliation

Rubrii's homepage says cases are not stored and remain on the user's device. The user guide repeats that patient data stays on the device. The privacy policy is broader. It lists search queries, repertorisation activity, case files, and patient information among information that may be collected, and says usage data can be retained for two years. It also describes account, payment, log, backup, analytics, and service-provider processing.

These statements are not necessarily impossible to reconcile. The policy may cover optional account features, support submissions, future services, or information that the current repertory flow does not transmit. It may also be a general policy that is broader than the implemented product. The public pages do not provide enough detail to determine the boundary confidently.

A clinic should ask Rubrii for a current data-flow explanation. Which fields stay only in browser storage? Are rubric searches or case-panel entries sent to a server? Do AI-powered searches transmit the full query? Are exported files uploaded before download? What does the service log? Which third parties receive data? Can an account owner export and delete account and usage data? How are backups handled? The answers should match a network test of the live app.

This is a procurement question, not a claim that the product mishandles data. The honest conclusion from the current pages is that the marketing and policy scopes differ and need written clarification.

Where Homeo AI fits differently

Rubrii starts with search and selection. Homeo AI starts with the narrative. A consultation can contain uncertain emphasis, negation, chronology, modalities, colloquial expressions, mental and physical qualifiers, and practitioner corrections. The review problem is preserving those details while proposing a structure that a qualified practitioner can change.

Homeo AI should therefore be evaluated on whether the original wording remains close to each proposed symptom and rubric, whether accepted and rejected mappings are visible, whether cited passages support the accepted interpretation, and whether follow-up changes preserve the earlier rationale. Its useful output is a reconstructable reasoning path, not simply a remedy ranking.

Homeo AI does not replace a complete repertory inventory or make Rubrii's matrix unnecessary. A clinic may prefer Rubrii when the practitioner already knows what to repertorise. A clinic may prefer Homeo AI when translating a long consultation into a reviewable totality is the larger burden. A combined process is possible, but only if rubric names, source editions, weights, and corrections survive the handoff.

Homeo AI also requires its own privacy review. Confirm current hosting, account access, processing, retention, deletion, export, support access, logs, subprocessors, and incident handling. No product should receive identifiable patient information merely because another product's privacy wording appears broader.

Source coverage and known limits

Rubrii's homepage advertises source badges and a large library, but the public pages reviewed here do not provide a complete edition-by-edition manifest for every repertory and materia medica source. Before relying on a result, request the exact title, author, edition, year, licensing status, update process, and mapping used for the tested content. Compare source badges with the passages and rubric hierarchy in the clinic's trusted reference.

AI-powered or synonym search can improve recall, but it can also surface a nearby rubric that changes meaning. A matrix can calculate consistently while still reflecting an incorrect input. A local case can reduce one privacy risk while leaving account, search, analytics, export, device, or backup risks. Each control solves a particular problem and should not be treated as a general quality guarantee.

Homeo AI has parallel limits. Proposed symptoms and mappings may be wrong, sources may be incomplete, and generated explanations may sound more certain than the evidence. Reviewability helps only when the practitioner actually inspects and corrects the output.

A reproducible clinic test

Prepare twelve completed, de-identified cases with a separate answer packet. Include short acute histories, long chronic narratives, follow-ups, negated symptoms, regional expressions, conflicting modalities, and cases where two plausible rubric choices lead to different remedy coverage. Remove names, contact details, dates of birth, exact dates, addresses, record numbers, images, and rare combinations that could identify anyone.

For Rubrii, begin with six cases whose accepted rubrics are already known. Search each rubric using the wording in the case and a synonym. Record the source badge, hierarchy, search precision, false-near matches, time to select the correct rubric, weighting behavior, matrix output, materia medica handoff, and PDF export. Compare the output against the trusted reference edition in the answer packet.

For Homeo AI, begin from the raw de-identified narratives. Record how much original wording survives structuring, which proposed symptoms and mappings are accepted or rejected, whether sources support the accepted interpretation, correction time, differential clarity, and follow-up continuity. Then compare the final accepted rubric set with the one used in Rubrii.

Run a separate privacy test with synthetic data that contains obvious markers but no real patient information. Use browser developer tools or an approved network monitor to observe what leaves the device during Rubrii search, case building, AI search, export, login, and deletion. Clear browser storage and repeat. Ask the vendor to explain any observed requests. Perform an equivalent product-specific data-flow test for Homeo AI.

Have a second qualified practitioner review the blinded exports and reasoning trails. Score rubric accuracy, source traceability, narrative fidelity, correction burden, reproducibility, and privacy clarity separately. Do not let a high remedy overlap hide a wrong rubric, and do not let an attractive explanation hide a missing source.

Use only completed, de-identified cases until the clinic has approved the full data flow and legal basis. Neither tool replaces diagnosis, referral, emergency care, practitioner judgment, prescription, potency, dosage, or follow-up responsibility. Neither tool should make an emergency-care decision, issue treatment instructions to consumers, or decide whether urgent medical assessment is required.

Conditional verdict

Choose Rubrii when the practitioner already has a sound totality and wants a focused, browser-based path through rubric search, weighting, repertorisation, materia medica, and export. Verify its source manifest and reconcile the current on-device claims with the privacy policy before using identifiable information.

Choose Homeo AI when the harder job is preserving consultation language, reviewing proposed symptom structure and rubric mappings, checking source passages, documenting a differential, and carrying the reasoning into follow-ups.

Use both only with a controlled handoff. Keep the original narrative and practitioner corrections visible, record the source edition behind each accepted rubric, and verify that no local or cloud step silently changes the case. The better workflow is the one that lets another qualified practitioner reconstruct and challenge the reasoning.

Official sources checked

• Rubrii official product page
• Rubrii official user guide
• Rubrii Privacy Policy, updated January 2025
• Rubrii Terms of Service, updated January 2025

Product facts reviewed 2026-07-28. Confirm current features, prices, source libraries, and terms with the vendor.