No telemetry
No data, profile or finding is sent to us, and no usage statistics either. Scanning needs no internet connection.
Security
Built first for teams whose data can't go to an outside service: banks, insurers, healthcare and public bodies.
No data, profile or finding is sent to us. No telemetry.
Read-only, in layers
Every connection gets the strongest read-only mode its database offers, and every query passes a guard before it runs. Datainspec records which protection was in effect for each connection.
Whatever the database, connect with a user that can only SELECT. The layers above are there for when that wasn't possible.
What stays on your machine
No data, profile or finding is sent to us, and no usage statistics either. Scanning needs no internet connection.
By default each finding is explained from a template. If you install a local model, it sees only the finding's summary, never a row of data, and every column and number it mentions is checked against the finding. If it invents one, it is asked again, and then the template is used.
Profiles, approved versions, decisions and reports live in Datainspec's working folder on your machine, separate from your sources.
Sharing and moving
Counts, rates and test statistics, with every id and value withheld or replaced by a digest made with your own key, so the same value gives the same digest across your reports. Notes and reasons people typed are left out by default, because they can quote data values. Column, dataset and reviewer names stay.
A bundle carries one dataset's approved versions, decisions and, if you choose, its snapshots, signed with an Ed25519 key. The import trusts only the public key you give it, never one the bundle names; a dry run shows what it would change first; importing twice changes nothing, and two conflicting histories are refused, not merged silently.
For security and IT reviewers
No. Datainspec runs where your data is and sends nothing to us: no data, no profile, no finding, and no telemetry.
No. It only reads your sources, with read-only access enforced in layers and tested against hostile queries. Its own records and reports are stored separately.
The audit log keeps every approval, block, rollback, role correction and threshold override, with the person, the time and the reason. Filter it and export it as CSV. Each approved version also has its own approval certificate.
Yes. A share-safe export contains no values from the data: counts, rates and statistics only, with ids and values withheld.
Yes. Run it from the app, or from your own CI. Exit code 0 means clear, 1 warning, 2 blocked, 3 the scan failed. Datainspec gives the verdict; your pipeline decides what to do with it.
The checks are statistical and rule-based, and each finding shows its evidence. Explanations are written from templates, or by a local model on your machine if you install one.
A pilot installs on your machines. Nothing is sent to us.