This document explains how data is handled in the My Conversation Partner feature. It is SEPARATE from the LEARNOVA Privacy Policy that applies to the platform as a whole.
This document is a DRAFT. Because the feature is not active, it does not yet describe processing that actually happens — it describes a DESIGN and marks the parts still awaiting a decision.
1. Document status and feature status
This document is a DRAFT and is not in force.
My Conversation Partner is not active. At the time of writing, no user audio, transcript or conversation text is processed by this feature.
2. Data designed to be processed
If the feature is enabled, the following data is planned to be processed:
- The learner's speech recording while practising.
- The speech-recognition transcript of that recording.
- Conversation text (learner utterances and system replies).
- Facts the learner states during practice (e.g. name, city, job) used to keep the conversation coherent.
- Technical metadata required for the service to run.
3. Purpose of processing
Data is processed to run the conversation practice itself:
- converting speech to text so the system can understand it;
- composing a reply that fits the conversation context;
- keeping the conversation continuous within a session;
- calculating minute-quota usage.
Any use beyond the purposes above — including marketing — has not been decided: [DECISION REQUIRED BEFORE PUBLICATION].
4. Third-party service providers
This feature is designed to use third-party services for speech recognition (STT) and language processing (LLM).
The following is NOT YET DECIDED and is deliberately left blank rather than estimated:
- Final STT provider — [DECISION REQUIRED BEFORE PUBLICATION]
- Final LLM provider — [DECISION REQUIRED BEFORE PUBLICATION]
- Model used — [DECISION REQUIRED BEFORE PUBLICATION]
- Data processing region — [DECISION REQUIRED BEFORE PUBLICATION]
- Whether data leaves Indonesia — [DECISION REQUIRED BEFORE PUBLICATION]
5. Storage location and retention
Where audio and transcripts are stored, and how long each is kept, is NOT YET DECIDED: [DECISION REQUIRED BEFORE PUBLICATION].
Until that decision is made, this document does not claim that data is stored in a particular location, nor that it is not stored.
6. Use for model training
Whether conversation data is used to train or improve models is NOT YET DECIDED and not yet proven: [DECISION REQUIRED BEFORE PUBLICATION].
This document makes no claim that data is not used for training until that is proven on the configuration actually used.
7. Data deletion
The mechanism for deleting conversation data, how long it takes, and whether users can request deletion themselves is NOT YET DECIDED: [DECISION REQUIRED BEFORE PUBLICATION].
8. Security
Access to conversation data is limited to the systems that process it and to authorised administrators.
Technical security controls for this feature will be stated once the implementation is decided: [DECISION REQUIRED BEFORE PUBLICATION].
9. Your rights
Your personal-data rights on the LEARNOVA platform are set out in the LEARNOVA Privacy Policy in force.
Any additional right specific to conversation data will be stated here before the feature is enabled: [DECISION REQUIRED BEFORE PUBLICATION].
10. Minors
Age requirements follow the LEARNOVA Terms of Service in force.
Any provision specific to voice conversation for minors will be stated before the feature is enabled: [DECISION REQUIRED BEFORE PUBLICATION].
11. Changes to this policy
This policy will be updated before the feature becomes available to the general public.
Every part still marked [DECISION REQUIRED BEFORE PUBLICATION] must be resolved first.
12. Contact
Questions about this draft document: support@learnova.fun.
The LEARNOVA Privacy Policy in force: https://learnova.fun.