Data residency

Data residency answers two questions your legal and security teams will ask: where is our data kept, and where is it processed? The Data Residency tab controls both. Storage is in the European Union by default. On top of that, you can require EU-served inference endpoints or zero-data-retention providers, and see live which models remain available before you commit.

app.stickyprompts.com/workspace#data-residency 1 EU by default 2 EU-served only 3 No retention by provider

Storage region

Storage Region is where your workspace’s files and prompt data are stored.

Region	Availability
European Union	The default. Chats, files and audit records are stored and processed inside the EU.
United States	On request, on Enterprise plans. The StickyPrompts team provisions US workspaces after a short residency review. Contact sales@stickyprompts.com.

You also meet this choice when you sign up: the Create an account page has a Data residency choice with European Union as the default and United States on request.

app.stickyprompts.com/registration

Inference constraints

Inference Constraints restrict which AI providers your workspace’s chats, prompts and agents may use. Each is a switch.

Constraint	Effect
Require EU-served inference endpoints	Only providers with an EU-hosted endpoint may run this workspace’s chats, prompts and agents.
Zero data retention only	Only providers that contractually do not retain your data may be used.

You can turn on either or both.

Origin is not location

Require EU-served inference endpoints is about where a model is served, not who made it. A model built by a US or Asian company can still run on an endpoint hosted in the EU, and then it stays available under the constraint. A model with no EU-served endpoint is removed, whoever made it. So the constraint does not shrink your choice to European model makers - it keeps processing in the EU while leaving you as much choice of model as the EU endpoints allow.

See the impact before you save

Impact on your model catalog shows what happens to your models under the constraints you have selected, updated live as you toggle them - for example Models usable 86 / 86 with no constraints on. Toggle a constraint, read the impact, and only then keep it. That way nobody finds out on Monday morning that their favourite model has gone.

1. Open Data Residency

   On the Workspace page, select Data Residency.

2. Try a constraint

   Switch on Require EU-served inference endpoints, Zero data retention only, or both.

3. Read the impact

   Check Impact on your model catalog: how many models stay usable. If a model a team depends on would go, talk to that team first.

4. Keep it or switch it back

   Keep the constraint if the result works for your company, or switch it off again.

How it interacts with model access

Data residency comes first. Models it rules out are greyed out on Model & Provider Access and cannot be used under any circumstances - not by members, not by specific teams, and not by admins. Model & Provider Access then decides who can use the models that remain.

	Data residency	Model & provider access
Decides	Which models can run at all	Who can use them
Based on	Where the model is served, and provider data retention	Provider, category, model and team
Admins exempt?	No	Yes, admins can always use every allowed model

Related