Ephemeral vs
Persistent Scanning

Choose the scanning protocol that aligns with your security requirements and operational needs. Ephemeral scanning provides maximum security through zero-knowledge processing, while persistent scanning enables deeper analysis capabilities.

Feature

Ephemeral

Zero-Knowledge

Persistent

Full Analysis

Source Storage

How source code is handled during and after analysis

PURGED

Source code deleted immediately after analysis. Only encrypted intelligence and findings retained.

STORED

Source code archived in encrypted S3 storage for future deep analysis and semantic queries.

AI Search

Semantic search and code intelligence capabilities

ARCHIVED

Search operates on encrypted intelligence archive. Limited to architectural blueprints and key code snippets.

FULL

Complete codebase available for semantic search, pattern analysis, and deep code intelligence queries.

Security Level

Data protection and attack surface reduction

MAXIMUM

Zero-knowledge processing eliminates persistent data storage. Minimal attack surface with immediate cleanup.

ENHANCED

Encrypted storage with access controls. Larger attack surface due to persistent data retention.

Processing Speed

Analysis completion time and resource utilization

FAST

Streamlined processing with immediate cleanup. Optimized for speed and minimal resource usage.

DEEP

Comprehensive analysis with full codebase indexing. Slower processing but deeper intelligence gathering.

Compliance

Regulatory compliance and data protection standards

GDPR/SOC2

Data minimization compliant. No persistent storage of sensitive source code. Full audit trails maintained.

ENHANCED

Encrypted storage with compliance controls. Additional security measures required for sensitive data.

[01] Choose Ephemeral When

  • Maximum security and data minimization are required
  • Source code contains sensitive intellectual property
  • Regulatory compliance requires zero persistent storage
  • Fast processing with basic rule-based analysis suffices
  • Cost optimization through minimal storage requirements

[02] Choose Persistent When

  • Deep semantic search and AI analysis are required
  • Historical codebase evolution tracking is needed
  • Complex refactoring and architectural analysis required
  • Team collaboration with shared codebase access needed
  • Advanced AI features like model fine-tuning desired

Ephemeral Processing Phases

Phase 1: Scanning

Code is analyzed in isolated memory buffers. Rules are evaluated against the codebase. Intelligence extraction begins.

Duration: 30-120 seconds

Phase 2: Persistence

Findings, code snippets, and architectural intelligence are encrypted and stored. AI context is archived for future queries.

Duration: 5-15 seconds

Phase 3: Cleanup

All source code artifacts are permanently deleted. Memory buffers are cleared. Only encrypted intelligence remains.

Duration: < 5 seconds