Arize AI

Model Monitoring

Production monitoring for performance, drift, data quality, and fairness regressions.

LLM Observability

Logging, tracing, and metrics for LLM applications in production — latency, cost, quality, and error rates across traces and users.

Agent Tracing

End-to-end visibility into multi-step LLM agent runs: tool calls, intermediate reasoning, token usage, latency, and errors at each step.

Drift Detection

Automated detection of distribution shift, feature drift, prediction drift, and performance degradation in deployed ML/AI models.

LLM Evaluation

Systematic testing of LLM outputs for correctness, relevance, safety, and consistency using automated scorers, rubrics, or human review.

Prompt Management

Versioning, templating, A/B testing, and deployment workflows for LLM prompts treated as production artifacts.

Bias & Fairness Testing

Automated statistical testing for disparate impact across protected attributes, with audit-ready reports.

Audit Logging

Tamper-evident logging of governance events (approvals, model changes, policy decisions) required by EU AI Act Article 12 and similar regulations.

Hallucination Detection

Automated scoring of LLM outputs for factual accuracy and groundedness, typically via retrieval-backed verification or reference comparison.

LLM Guardrails

Runtime input/output filtering for LLMs — PII redaction, toxicity blocking, prompt injection defense, policy enforcement.