Files to forms. Forms to files.

Entruno privacy policy

Effective 16 October 2026. Contact: support@brilliantmindsglobal.com.

Entruno helps a user review source data, fill supported website forms, and export records from an explicitly selected website. This policy describes the 0.9.3 candidate, including automatic upload and multi-record document reading.

Website access and data on your device

Entruno requests permission for websites you approve. It reads supported field labels, control structure, the selected form or record list, and the values you choose to process. Uploads and submissions use your existing website session. In the simple upload screen, choosing a file under Upload & submit initiates automatic extraction, matching and sequential submission. The first saved record is verified before the batch continues. Missing or uncertain information stops the job. Advanced tools retain an explicit review workflow. The extension is bound to the originating tab. It does not read Chrome's browsing-history database or track unrelated tabs. It retains the URLs, page context and job checkpoints for website tasks that you initiate, including URLs in saved website setups and collected record links. It records job progress and outcomes for recovery, rather than general click or keystroke monitoring.

Excel rows, selected attachments, document drafts, job outcomes, downloaded records and bounded attachment-download caches are stored in your Chrome profile. Website setup memory contains the URL and structural mapping, rather than record values or file contents. A matching URL and schema can reuse setup without another AI request. A changed schema is inspected again and uncertain matches need your input. Profile storage can be inspected by someone with access to that browser or operating-system account. Per-account separation in the interface is not encryption or protection against the owner of the computer. Use separate Chrome/OS profiles on shared computers.

Source documents and chosen website fields can contain personal identifiers, health or financial information, personal communications and location information. Entruno processes the content you select for the stated upload/download task; a general document may contain several of these categories. Only choose material that you are authorised to process and, for cloud reading, send to the configured services.

AI and service providers

Ordinary AI field matching sends column names and website field labels/structure to the configured team service, which calls OpenAI. Record values and attachment bytes are not included in this ordinary matching request. A label itself may contain sensitive text; review your source labels before enabling AI. Export planning sends table headings, identifier/detail-link presence and page-control labels, not the collected record values.

In the simple upload screen, choosing a PDF or photo under Upload & submit authorises cloud reading as explained beside the file chooser. Up to 20 rendered pages, including all visible values and portraits on those pages, and target form labels are transmitted over HTTPS to the configured team service and OpenAI. There is no additional per-document modal in this screen. Multiple register rows can be extracted and matched automatically; uncertainty or an incomplete extraction stops submission for your input. Embedded supporting files are extracted locally and sent only to the chosen destination website when required by the form. Image-only exact-file document drafts are remembered locally to avoid repeated AI calls; the cache stores extracted values, not rendered page images. Advanced tools still offer local reading and the older per-document cloud consent flow, including a separate consent prompt. Choose only documents you are authorised to send to these services.

The team API key is stored only on the team server in a restricted-permission database. Team members receive their own expiring access tokens, not that API key. API requests use OpenAI's store:false setting. That setting does not mean zero provider retention; OpenAI's applicable API data controls and contractual retention apply. See https://platform.openai.com/docs/guides/your-data . AWS hosts the team service and encrypted recovery backups. Your configured destination website receives any form data or files you approve for submission. Entruno does not control that website's retention or access rules.

Retention and deletion

Local upload/download jobs and saved job snapshots remain until replaced or deleted. The data-management screen can delete the current Entruno account's local upload/download data. Website setup memory is account-separated and bounded to 200 entries and a 4 MiB budget across the Chrome profile; it is retained until forgotten, evicted, or deleted. Cloud document drafts expire logically after 30 days and are capped at 20 per account; expiry is checked on reuse and cleanup occurs during later reads. Browser uninstall/profile removal also removes extension storage, subject to Chrome and operating-system behaviour. Deleting extension data does not delete files already downloaded to disk or records already submitted to a website.

The team server stores usernames, salted password hashes, hashed session tokens, the administrator's provider key, fixed event identifiers, quota counters, bounded AI setup cache, and numeric provider usage/latency measurements. Sessions expire after 30 days and can be revoked. Setup responses expire after seven days and are capped at 500. Usage measurements are capped at 5,000 entries/30 days, with cleanup during requests; fixed event logs retain the last 5,000 events. The service does not persist document images or extracted document values. Infrastructure access logs may contain request paths, times and network identifiers; do not put secrets in URLs. Administrators handle account/server-data deletion requests through the support contact.

Encrypted off-host database backups have a 30-day object lifecycle and seven-day noncurrent-version lifecycle; deletion is subject to AWS lifecycle processing. Recovery keys are kept separately in AWS Systems Manager Parameter Store SecureString. Backups contain account configuration and the provider key, so recovery access is restricted. Scheduled backup operation is a release gate; a successful manual restore drill does not establish an ongoing recovery service.

Exports and safeguards

Excel/PDF exports contain the selected website's records and permitted photo/document URLs. The optional bounded ZIP export includes same-origin attachment bytes and a checksum manifest. Signed URLs may expire and URL query strings can be sensitive. Downloaded exports and local data-backup files are your responsibility. Local data backups are explicitly unencrypted and contain records/files; they exclude login tokens and the provider key. Protect them and delete them when no longer needed.

Entruno has no advertising or sale of user data. Data is used to deliver these user-requested features and operate the team service. Use of data received through Chrome APIs complies with the Chrome Web Store User Data Policy, including its Limited Use requirements. No model output is executed as code. AI may be wrong. Automatic mode stops on reported uncertainty, missing required data and unsupported controls, and requires a fresh identity-matching receipt before counting a save. Operator correction remains available for exceptions. CAPTCHA, OTP and inaccessible or unsupported controls need manual handling.

Contact and changes

For support, deletion requests or questions, email support@brilliantmindsglobal.com. Material changes to data handling will be reflected in this policy and the product's file-selection notice and applicable consent flow before the changed feature is used.