Changelog

What's New

Track our journey from launch to the future of AI safety infrastructure.

v1.0.0 Q1 2026 Launch

DriftRail Launch

The complete AI safety and observability platform. Monitor every LLM interaction, catch risks before they reach users, and stay compliant without slowing down.

AI-Powered Risk Detection

  • + Hallucination detection with source verification
  • + PII exposure scanning (GDPR/CCPA compliant)
  • + Prompt injection and jailbreak detection
  • + Toxicity and harmful content filtering
  • + Custom detection rules — define your own classifiers

Real-time Guardrails API

  • + Sub-50ms inline content checks
  • + Block, flag, redact, or warn actions
  • + Automatic PII redaction before storage
  • + Fail-open design — never breaks your app

Full-Stack Observability

  • + Real-time event streaming dashboard
  • + Model comparison and A/B testing
  • + Latency, cost, and error rate tracking
  • + Executive dashboards with trend analysis

Enterprise Ready

  • + HIPAA, SOC2, GDPR, PCI-DSS compliance
  • + Incident management with MTTR tracking
  • + Immutable audit logs with tamper detection
  • + SIEM integration (Splunk, Datadog)

Developer Experience

  • + Python and Node.js SDKs with TypeScript
  • + AI Playground for testing with live detection
  • + Scoped API keys with IP allowlisting
  • + Slack, Teams, Discord integrations
v1.1.0 Q2 2026 Roadmap

On the Horizon

  • ○ OpenTelemetry native integration
  • ○ LangChain and LlamaIndex auto-instrumentation
  • ○ Custom classifier fine-tuning
  • ○ Real-time streaming log viewer
  • ○ Multi-tenant SSO and RBAC