Updated November 6, 2026
This policy covers Ullman Indirect for Android (package com.ullman.ullmanindirect). Contact mullman@mycallbag.com with privacy or support questions.
Ullman Indirect uses the camera and images or videos you select to capture, crop, label, draw on, and combine retinal images. Image editing, local image analysis and montage processing run on your device. The app has no developer-hosted patient account, patient cloud database, advertising service or in-app purchase system.
Optional patient name, date of birth, medical record number, notes and date/time preferences are held for the current app session. They may be written into a photo, drawing or other result that you save or share. Review each result before exporting it. Camera and onboarding preferences are stored privately on your device. Unsaved patient fields and editing state are not a permanent patient record and can be lost when the process closes.
Scan Patient Details uses bundled Google ML Kit text recognition. Its image and recognized text are processed on-device, and you can review the text before adding it to notes. Face detection used to help position an image also runs on-device. ML Kit does not send these input images or recognition results to Google for processing. Its separate technical metrics are described below.
Dictate Patient Details requests microphone permission only when you choose to record. It uses your device’s speech recognition provider. The app requests a preference for offline recognition, but the provider may ignore that preference and send audio to its servers. Spoken patient details can therefore be processed by that provider. Its privacy practices, transport and retention depend on the service and your device/account settings. You can type notes instead. The app does not save a separate dictation audio file; review the returned transcript before adding it to notes. Camera videos are recorded without a microphone audio track in this version.
Google ML Kit may send device and app information, installation identifiers, feature events, errors and performance measurements to Google for diagnostics and usage analytics. Google documents HTTPS transport for these metrics and no onward transfer to third parties. This technical collection is separate from the clinical images and recognized text processed locally.
Optional montage models are downloaded through Google Play Asset Delivery when needed. Google processes device metadata and app version to deliver and maintain the appropriate assets. Google documents encrypted transport, no transfer to third parties, and a fixed retention period for this delivery data; no specific deletion deadline is promised here.
Montage inference uses ONNX Runtime locally. This app sets its telemetry-disable configuration before initialization and also disables telemetry through the runtime API. Images are not sent to Microsoft for inference. These settings do not disable the separate Google services or your selected speech recognition provider.
Captures and working files can be held in private app memory, cache or capture storage while you work. Saved images are written to your gallery, normally Pictures/Ullman Indirect, and saved videos to Movies/Ullman Indirect on newer Android versions. Depending on the feature and selected settings, saved images use JPEG or PNG. You decide whether to use Android’s sharing chooser and which destination receives the selected file. Sharing grants that destination access to the file, including any patient information rendered into it.
Gallery, messaging, email and cloud-backup services you choose have their own practices. The app does not automatically upload your clinical results to a developer server. Optional dictation and technical SDK data are the exceptions to an entirely local workflow described above.
You can review or revoke camera and microphone permissions in Android Settings. Available controls and speech services vary by device.
Use Clear in Patient Info to clear the current draft and Save to apply it. Clear app storage in Android Settings or uninstall the app to remove private app files and preferences. Temporary files can be cleaned up by the app or Android; no fixed lifetime is promised for every working file. Successfully shared temporary exports can remain in the app’s cache for later access and are eligible for cleanup during a subsequent export after 24 hours.
The app disables its private Android backup and excludes private app storage from cloud backup and device transfer through its backup configuration. Public gallery files and copies held by other apps are outside that exclusion. Delete saved images and videos separately in your gallery, including any Trash or Recently Deleted folder. Remove shared/cloud copies through their respective services. Clearing this app’s storage or uninstalling it does not erase records retained by Google, a speech provider or a sharing recipient.
There is no app account to delete and no developer-held patient database. Contact support for questions about the app’s handling of data. Do not send patient images, identifying details or medical records in a support message unless a suitable authorized process has been agreed.
Use a secured device and obtain appropriate authorization before capturing or sharing another person’s information. We do not promise that a gallery, speech service or sharing destination meets a particular clinical privacy standard. No image-quality or diagnostic-accuracy guarantee is made. Seek appropriate professional care for medical concerns.
We may update this policy when the app or its services change. The date above identifies the current version.