Medil AIResearch record
The September 2026 preprint describes the design and implemented functions of Medil AI version 4.17 on Android and iOS. This page makes its methods, dated findings, evidence locations, disclosures, and reusable citation formats easy to inspect.
Ariel Israel and Eugene Merzon
Preprints.org, version 1, posted 7 September 2026. DOI 10.20944/preprints202609.0448.v1.
The article is available under CC BY 4.0. Cite the authors and preprint when reusing its material.
The paper is a descriptive design and implementation study of the patient-facing workflow, its information sources, and its confirmation steps. The authors mapped needs identified from clinical experience and literature on medication literacy, adherence, language access, and patient-generated health data to implemented app functions.
The documented evidence came from released app screens and public patient and clinician guides for version 4.17. Availability was classified by platform, market, and subscription status. Reporting was informed by the mERA checklist and the iCHECK-DH guideline.
As documented in the September 2026 preprint, Medil AI version 4.17 connected source-linked medicine information, prescription capture, personal instructions, reminders, patient-confirmed dose records, prescribed dose-sequence support, and a medical diary across Android and iOS.
The implementation snapshot reported medicine information drawn from public regulatory or authoritative sources in 12 country environments and translation into 20 languages. It also documented authorized Health Connect and HealthKit imports and camera-assisted transcription of values shown on separate home-monitoring devices, with review before saving.
This summary describes the version 4.17 research snapshot. See current app facts for later Android and iOS releases.
The preprint reports 12 country-specific data environments, with an additional Europe-wide authorization layer for other European countries. Available fields and update timing follow each public source.
| Paper reference | Topic | What it documents |
|---|---|---|
| Table 1 (pages 4-5) | Post-consultation needs mapped to app functions | Links medicine understanding, personal instructions, official notices, language access, prescription capture, reminders, changing-dose schedules, diary records, camera entry, and visit preparation to a corresponding control or qualification. |
| Table 2 (pages 6-7) | Provenance and confirmation | Distinguishes source-linked information, generated summaries, image-read proposals, reminders, dose-sequence templates, self-reported doses, health readings, the diary, and external-AI handoffs by origin and required interpretation. |
| Table 3 (pages 7-8) | Medicine-data environments | Names the public regulatory or authoritative source used for 12 country-specific medicine-data environments and an additional Europe-wide authorization layer. |
| Figure 1 (page 9) | The connected patient workflow | Maps medicine understanding, language access, a confirmed plan, reminders, patient-reported dose records, and a diary into one post-consultation path. |
| Table 4 (pages 9-11) | Implemented functions and access status | Records the version 4.17 feature inventory and whether each function was available without a subscription or through Medil AI+ after a free trial. |
| Figure 2 (page 12) | Prescription-image workflow | Shows framing, editable proposed medicine and instruction fields, a low-confidence field flagged for review, and the resulting patient-confirmed care plan. |
| Figure 3 (page 13) | Reminder, dose record, and external-AI handoff | Documents a reminder action, full or partial self-reported dose recording with notes and undo, care events, and user selection of context before leaving the app for an external assistant. |
| Figure 4 (page 14) | Language and cross-script access | Shows source-linked medicine information, on-demand translation, right-to-left presentation, Cyrillic search, transliteration, and a national medicine record. |
| Figure 5 (page 15) | Prescribed dose-sequence confirmation | Shows a product-specific sequence with each phase, dose, and duration presented before the user confirms that it matches the prescriber's instructions. |
| Figure 6 (pages 16-17) | Camera-assisted reading and the combined diary | Shows proposed values read from a separate monitor display for review or correction, and a month view combining medication-use records with selected patient-generated measurements and events. |
Table 2 treats provenance and confirmation as part of the function, not as fine print. The distinctions below are paraphrased from that table.
| Element | Role in the app | Control or interpretation |
|---|---|---|
| Source-linked medicine information | Public regulatory or authoritative medicine-data source | The source and its update timing govern the information shown. |
| Plain-language summary or translation | Generated reading aid associated with source text | The source remains available beside the reading aid. |
| Prescription read from an image | Editable transcription proposal | The user compares and confirms every field before saving. |
| Medication reminder | Prompt for a time in the confirmed plan | Delivery depends on the phone's notification and power settings. |
| Prescribed dose-sequence template | Product-specific phases presented before scheduling | The user confirms the sequence against prescriber instructions. |
| Self-reported dose record | Dated record of a full, partial, or not-taken dose | It shows the user's own confirmation and any note on the timeline. |
| Health reading | Manual entry, authorized health-platform import, or display reading | It keeps the value, source, and time together as a diary entry. |
| Camera-assisted measurement entry | Transcription of visible values from a separate device display | The user reviews or corrects the proposed values before saving. |
| Diary and chart | Dated medication, event, note, and measurement record | It organizes what happened and when for later review. |
| Question sent to an external assistant | Question plus categories of personal context selected by the user | The conversation takes place in the selected external service. |
This is the access model reported in Table 4 for the 3 September 2026 implementation snapshot. Current Android 4.20 and iOS 4.20 releases require Medil AI+ for the entire Medical diary, including manual measurement entry, calendar and charts; these features are included during an active eligible trial. See current app facts for the complete current access model.
The report used software, public product documentation, demonstration records, and application screens. It involved no research participants or identifiable patient records, and no participant dataset was used.
Ariel Israel designed the application, formulated the study and manuscript concept, assembled the documentation, and drafted and revised the manuscript. Eugene Merzon contributed to study design, drafting, and critical revision.
The authors reported no financial support for the research, authorship, or publication. Ariel Israel disclosed that he is the creator of Medil AI and founder CEO of Medil Health Technologies Ltd and may benefit financially from adoption; Eugene Merzon declared no competing interest.
The authors reported using ChatGPT and Claude for manuscript editing, linguistic clarity, and reference search. They reviewed the assisted revisions and accepted responsibility for the manuscript.
Use the factual media kit, read the medicine-source methodology, or inspect the illustrated patient guide. The official app is available on Google Play and the App Store.
Questions, corrections, or requests for supporting material: support@medil.ai.
We use consent-gated analytics to understand how people find Medil AI. We do not collect medication names, searches or health data on this site. Privacy policy