Security & Privacy

Security and data protection are built into Replion from day one - not added as an afterthought.
We help organizations use AI responsibly, securely, and with full control over their data.

Quick facts

• Data hosted in Sweden / EU
• Tenant isolation (customer data is separated by design)
• Encryption in transit and at rest
• Default retention: 90 days (configurable)
• DPA available/included for customers

Your Data, Your Control

Replion is designed so your data remains under your control.

Key points

• Customer documents and knowledge are isolated per account (tenant)
• Data is not shared across customers
• Customer content is processed only to provide the Service
• Customer content is not used to train Replion’s own models
• Documents and conversations can be exported and deleted (based on your retentionsettings)

Secure by Design

We apply modern security practices across infrastructure and application layers.

Key points

• Encryption in transit (TLS) and at rest
• Access controls and least-privilege by default
• Secure authentication and protected APIs
• Monitoring of system health and abnormal activity
• Regular updates and hardening practices

EU & GDPR-Friendly by Default

Replion is built for organizations operating under European privacy requirements.

Key points

• Data hosted within Sweden / EU
• Supports GDPR compliance (controller/processor setup)
• Clear retention and deletion options (default 90 days)
• Data Processing Agreement (DPA) available for customers
• A current list of subprocessors is available on request - please email: support@replion.se

AI You Can Trust

Replion is designed to provide answers grounded in your approved sources.

Key points:

• Answers are generated from your connected content (PDFs, URLs, uploaded docs)
• If Replion can’t find a reliable answer, it falls back and/or escalates
• Escalations can include context (e.g., chat transcript) to help your team respond faster

Operational Transparency

Teams have visibility into what the system is doing, so you can continuously improve quality

Inside the dashboard, you can see:

• Which sources are being used
• When fallbacks/escalations happen
• High-level quality indicators (e.g., “answer confidence” / “source coverage”)

Responsible Incident Handling

Security is an ongoing process. If an incident occurs, we follow internal routines to:

• Detect and investigate abnormal behavior
• Limit impact through isolation and access controls
• • Notify customers when required and share relevant details