Skip to main content

Zero-Storage Principle

Superatom NEVER stores your business data. All processing is ephemeral.

What IS Stored

What is NOT Stored

  • Raw business data
  • Query results
  • Exported files
  • Customer PII from sources

Encryption

Data Lifecycle

1

Query Received

User asks a question
2

Data Retrieved

Query executes against source
3

Processing

AI analyzes results in memory
4

Response Sent

Visualization streamed to user
5

Cleanup

All temporary data deleted

What the LLM Receives

The LLM is called as a stateless inference service. Each request is constructed from scratch, and no state persists between calls.

Sent to the LLM

Never Sent to the LLM

Customer data, queries, and knowledge nodes are never used to train or fine-tune AI models. The LLM provider receives stateless inference requests only.

Air-Gapped Deployment

For environments that cannot make any external calls, Superatom supports self-hosted LLM deployment using open-source or licensed models running within the customer’s network. In this configuration:
  • Zero data leaves the network perimeter
  • All AI inference runs on-premise
  • Usage telemetry can be disabled
  • The platform operates completely independently

Network Security

  • Deploy within the customer’s network
  • No public database exposure
  • TLS 1.3 for all connections
  • No inbound ports required
  • Connection severable instantly
  • All outbound traffic can be proxied

Network & Data Flow

Full details on outbound traffic, what stays in-network, and air-gapped deployment