Privacy
Privacy Notice
Last updated: April 16, 2026
Liminaly is designed as a privacy-sensitive reflection space. We know people may use it to process personal, stressful, or high-stakes situations. Because of that, we try to minimize what we collect and keep product analytics separate from the private content you type.
What you write
When you use Liminaly, you may enter private text to receive a reflection. That text is processed to produce your output in the live flow and is not stored as saved history by the product.
We do not use the content of your private text entries or AI-generated reflections for analytics.
That means we do not use your raw reflections, generated summaries, threads, or next steps to measure product usage or behavior.
Minimal product telemetry
To improve reliability and understand whether the product flow is working, we may collect limited product telemetry.
This may include events such as:
- starting a reflection
- whether a reflection completed successfully
- whether a fallback result was used
- whether a user took a next step or marked it as done
- whether a shown reflection became stale before action
This telemetry is designed to be minimal and coarse. It does not include the content of what you wrote or the content of the result shown to you.
We also avoid invasive tracking methods. We do not use session replay, form capture, or content-based behavioral analytics in the standard product flow.
What telemetry may include
Our limited telemetry may include information such as:
- whether the result came from the main path or a simplified fallback
- broad language classification such as English or Indonesian
- coarse technical timing buckets
- whether a case was treated as high-stakes
- whether the app was used on mobile or desktop
- short-lived anonymous session information
This telemetry is used to understand product quality and reliability, not to profile your private situation.
What we do not collect for analytics
We do not use the following for analytics:
- raw text you type into Liminaly
- raw AI outputs
- generated summaries, threads, or next steps
- session replay recordings
- automatic form-content capture
- strong persistent identity for product telemetry
Optional feedback
If we ever add a way to report that a reflection missed the mark, any deeper example-sharing should be explicit and optional. The default product experience should remain private by design.
Data minimization
We try to collect only what is needed to operate the product, measure reliability, and improve the core flow. We try not to collect more simply because it would be useful.
Retention
We aim to keep raw telemetry limited and not retain it longer than necessary for product operations and analysis. Aggregated reporting may be retained longer than raw event-level telemetry.
Third parties
Where possible, we prefer first-party product telemetry rather than invasive third-party behavioral analytics. If we rely on third-party infrastructure or processors, they should be limited to what is necessary to run the service.
Your choices
The product is designed to stay light and disposable. You can use the live flow, act on the reflection, and move on without building an archive inside the app.
Changes
We may update this notice as the product evolves. If we make meaningful changes to how we collect or use data, we should update this page and revise the date above.
Contact
If you have questions about privacy in Liminaly, contact: [insert contact email]