Data retention policy
Kotobase only retains data as long as it is necessary to provide translation and collaboration services. By default, messages are processed in real time and are not stored permanently. Minimal metadata (such as user and workspace IDs) may be retained for service functionality, analytics, but no message content is stored beyond what is strictly required for translation.
Data archiving and removal policy
We do not archive user message content. Any cached data used for translation is automatically purged. Users may request removal of all associated metadata (e.g. workspace integration details, preferences), which will be permanently deleted from our systems.
Data storage policy
Kotobase stores only necessary configuration data (e.g., tokens, preferences, and workspace mapping) in encrypted databases. No user message content is stored long term. All data at rest is encrypted, and data in transit is protected.
Data center location(s)
Japan
Data hosting details
Cloud Hosted
App/service has sub-processors
no
App/service uses large language models (LLM)
yes
LLM model(s) used
OpenAI GPT-series models (via API)
LLM retention settings
We do not persist customer message content for LLM purposes.
Any data sent to the LLM is retained only temporarily for processing and is not stored by our application.
OpenAI’s API data retention policies apply.
LLM data tenancy policy
Customer data is processed on a per-request basis and is logically isolated per workspace.
Data sent to the LLM is used only to generate the requested translation or response and is not shared across customers.
We do not use customer data to train or fine
LLM data residency policy
LLM requests are processed via OpenAI’s API.
Data may be processed in regions where OpenAI operates its infrastructure, in accordance with OpenAI’s data residency and security policies.
We do not control or override OpenAI’s data processing locations.