Transparency: how your data is used

Last ​​updated: September 7, 2026 

Our Privacy Policy covers disclosures from a legal perspective. This page is different. It's the plain-English version — what we actually do, in practice, with what you put into Styleworks.

We think privacy is better explained by what we don't do than by what we're legally allowed to do. So that's where we'll start.

PerfectIt for Windows: no sensitive data leaves your PC

If you use PerfectIt for Windows, no data of any kind is sent anywhere, for any reason. PerfectIt for Windows runs entirely offline. Your document, and all other information, stays on your machine.

If you use PerfectIt Cloud then your data is on a server temporarily. It is encrypted in the send and on the server. It is not stored permanently. It’s quickly removed after you’re finished checking. We are also working on a new version that will keep all of your information on your machine.

We want to make sure that Styleworks reflects that same focus on data privacy and security.

Styleworks: what you give us

Styleworks is different. It’s an online service, so our servers do receive what you type into Styleworks when you build a style sheet or house style, that can include:

  • Terms to look for

  • Terms to suggest

  • Descriptions written for your team or clients

  • Terms that should be excluded

  • Any other detail you choose to add to your style sheet or house style

We want to be straightforward about that: yes, it is on our server which is hosted on Microsoft Azure (the server is in Northern Europe). Styleworks is a tool for managing your style sheets, house style and lexicons. So that data has to be online in order to run the product.

The information on Styleworks is just the data from your style sheet or house style. No information from the documents you check with PerfectIt is ever sent to Styleworks. Using Styleworks does not change how PerfectIt treats your sensitive data in any way.

What we never do with Styleworks data

Styleworks is not an AI product. The data that you enter is never fed into an AI model or automated training system of any kind.

We also never use any descriptions that you write, any exclusions you add, or any context you enter for your team. When you add that data to Styleworks, it sits in your account, for your use only.

Our team cannot see or read your style. Your data is private. The only way for our Support and Success teams to see your work is if you explicitly send it to them.

The one thing we do use (and why it's narrow on purpose)

The thing that makes Styleworks magical and unique is that you only have to enter the preferred version of a term and it knows the mistakes to look for. So, for example, you enter “email” and it can figure out that you want it to search for “e-mail” (with the hyphen). It can also figure out that the plural is implied from the singular.

However, that functionality is limited. It can only do that for the terms that we've given it data for. In order to make that feature truly magical, we need to add additional terms.

To ensure that we can continue to improve the product, we do use data on those pairings. We've made that approach as narrow as we can so that we don’t have access to anything sensitive. Here’s how it works.

We automatically create a report that is restricted in five ways:

1. Only two fields, ever. Of everything listed above, we only ever use the term to look for and the term to suggest. Nothing else — not your descriptions, not your exclusions, not any other detail in your style sheet. The report we use to improve the product only includes the two fields.

2. Only at export. The report only looks at term pairings at the point you export it to PerfectIt.

3. Only once it's common. A term pairing has to show up in three different style sheets before it makes it into the report. If it's specific to your organization, your client, or your project, it will never reach that threshold — and it never goes into the report.

4. Everything is anonymized. The report does not include names of users, emails, or any other personally identifying information.

5. A human checks it. There are no automated additions to the product. Every pairing in the report is checked personally by our analytical linguist. We screen out anything that looks like it could be confidential.

If you stop using Styleworks, there’s no way for us to undo that learning and make the product worse for everyone. So the narrow, anonymized data that makes it into that report is something we hold on to.

What this looks like once it's "in" the product

Even after a term pairing clears all of that, it's never shown to anyone. It doesn't appear in a list, a suggestion panel, or anywhere a user could browse it. The only effect is invisible: Styleworks gets a little better at recognizing that term and suggesting the right pairing. If (and only if) someone enters the term into Styleworks then Styleworks will suggest the right fix for all users.

The help that data provides

Styleworks improves for every user over time, using real patterns from real style sheets and house styles. We haven’t built an opt-out because we think it’s right that everyone who uses the product is making it a tiny bit better for everyone else. Styleworks achieves that without exposing anything confidential, and without needing anything beyond two fields ((with a cautious approach built in just in case they are sensitive).

If anything here raises a question, or you'd like more detail on any of this, email us at support@intelligentediting.com.