Personal data hunt
- Model
- Privacy FilterOpen
privacy-filterOpen model, not part of Orion. Finds and masks personal data in text.
A message full of made-up personal data. Mark what you find; the model looks for its own and you are both scored by category.
Live: this demo makes real calls to the API, with a demo key and a limit of calls per visit.
Select the words that are personal data, then press Check answers: one point per item found (marking half of its words is enough) and one point less per word marked by mistake. Then add some made-up details and see what the model finds.
Your marks
Hello, I am Lucy Fairweather and I am writing about my place on Saturday’s workshop. I was born on 14 March 1991 and I live at 7 Old Olive Street, flat 2A. You can call me on +44 7700 900117 or write to [email protected]. For the payment, my IBAN is GB00 ABCD 1234 5600 0000 00. My portfolio is at https://lucy-fairweather.example.org and the temporary password you gave me is Tq7pRw2zL. I will come with my colleague Ivo Saltmarsh.
What the model finds
Shown when you check
Up to 200 characters. Added text is not scored.
| Category | In the answer key | You | Model |
|---|---|---|---|
| Person | 2 | – | – |
| Address | 1 | – | – |
| 1 | – | – | |
| Phone | 1 | – | – |
| Date | 1 | – | – |
| Account number | 1 | – | – |
| URL | 1 | – | – |
| Secret | 1 | – | – |
Score
Answer, then press Check answers to see the model’s answers and the score.
Summary
- Model calls
- Median per call
- Cost at scale
With the local simulation, no cost or milliseconds.
Next: Exact clause searchModel calls
The cost and the milliseconds appear with the first call.
Milliseconds of each call
The local simulation does not call the API, so there is no cost or milliseconds to measure.
See the call
One request to POST /v1/redact with the whole text in inputs, placeholders on and every category.
- Route
POST /v1/redact- Model
privacy-filter
The server adds the API key, which never reaches your browser.
Request
API response
Take it to your code
The same routes find the answer in your own documents or mask the personal data in a text before you process it.