# naoveil

> naoveil is a de-identification tool for tabular clinical research data, made by naosuu LLC. It classifies each column of a CSV, XLSX or Excel worksheet, requires a person to review every decision, transforms the data on the user's computer, keeps pseudonyms and date offsets consistent across study files, and exports a processing log. Status: development alpha; a controlled research pilot is in preparation. Pricing has not been announced.

## Who it's for

- Clinical research data managers who prepare study exports for sharing.
- Investigators and study teams who send data to collaborators, consortium partners, journals or repositories.
- Privacy and compliance reviewers who approve data requests and need to see how a dataset was prepared.

## Use cases

- Sending a patient-level dataset to an outside collaborator.
- Monthly exports from a long-running study, where each patient keeps the same code (for example Subject_0137) in every file.
- Preparing a de-identified copy for a publication or data repository.
- Answering a data-use or ethics review with a record of how the data was prepared.

## What it does

- Checks each column of a CSV or Excel file and suggests an action: remove, disguise, generalize or keep. It flags names, IDs, dates, places and health details.
- A person approves every change. Export is blocked until every column has a decision, and any sensitive detail that is kept needs a stated reason.
- Typical results: names replaced with [REDACTED], patient IDs replaced with study codes, birth dates reduced to the year, ages grouped into ranges (90+ for the oldest).
- Each patient's dates move forward or back by the same number of days, up to a year, so the time between that patient's events stays the same.
- An encrypted study mapping file, protected by a passphrase, keeps patient codes and date shifts the same across exports. Patient IDs must match exactly; losing the file or passphrase means later exports cannot reuse the codes.
- Every export includes a processing record: the source file, the naoveil version, what happened to each column, any changes a reviewer made, and the reasons given for kept details.

## Where it runs

- Excel add-in: works on the active worksheet in Excel for Mac, Windows and the web; writes a processed sheet and a log sheet and never changes the source sheet.
- Desktop app (Windows, macOS): CSV and the first worksheet of XLSX files; saves only where the user chooses.
- Command line: the same engine from a configuration file.

## What naoveil does not do

- naoveil does not certify anonymization or HIPAA compliance. It supports a de-identification review; the determination belongs to the user's IRB, privacy office or a qualified expert.
- Classification is heuristic and has no accepted accuracy benchmark yet.
- Pseudonymized data can still be personal data.
- It works column by column and does not find identifiers inside free-text notes.
- The whole dataset is held in memory.

## Links

- [Product page](https://naoveil.naosuu.com/): methods, interactive figures, testing status, limitations and FAQ
- [Questions and answers](https://naoveil.naosuu.com/#faq)
- [naosuu](https://naosuu.com/): the company
- [naoveil privacy notice](https://naoveil.naosuu.com/privacy)
- [naoveil terms of use](https://naoveil.naosuu.com/terms)

## Contact

info@naosuu.com · naosuu LLC, 418 Broadway, Suite N, Albany, NY 12207, United States