Document
Trust & Security
Self-assessment
Self-assessment checklist (16 items)Privacy impact assessment (KISA PIA)
Personal data handling
Disclosure, processing, cross-border transfer
Security
Security and safeguards

Repattern
최종 업데이트: 2026-03-30

Disclosure, processing, cross-border transfer


Services used in production

AI model providers

VendorCountryData sentNot used for trainingAPI loggingZDR
OpenAIUnited StatesPrompts, conversation context✓ (store: false applied)OFFApplying
AnthropicUnited StatesPrompts, conversation context✓ (feedback sending OFF)OFFApplying
Google (Gemini)United StatesPrompts✓ (paid API baseline policy)OFFApplying

No AI provider uses customer data for model training, and API logging is disabled at all of them.
Details: Trust & Security

Document processing

VendorCountryData sentPurpose
Upstage (Document Parser)KoreaDocument filesDocument parsing and text extraction

Web search

VendorCountryData sentNotes
PerplexityUnited StatesSearch queriesDefault search provider in production

Infrastructure and cloud

VendorRegionUse
AWS S3SeoulFile storage
AWS RDSSeoulDatabase
Self-operatedSeoulSession and cache management

All AWS infrastructure runs in the Korea (Seoul) region.

CRM

VendorCountryData sentPurpose
HubSpotUnited StatesName, email address, company nameCRM / lead management

Consistency with the processing-policy vendor list

ItemStatus
AWS (cloud)Consistent - listed in both the processing and cross-border transfer tables
HubSpot (CRM)Consistent - listed in both the processing and cross-border transfer tables
OpenAI, Anthropic, Google, PerplexityConsistent - listed separately in the Repattern section

Every external service used in production is reflected in the privacy policy.