Monitoring platform spanning traditional ML models, LLM applications and multi-agent systems. For LLM applications customers publish prompts, prompt context, responses and retrieved source documents; Fiddler generates trust and safety metrics, embeddings with UMAP visualisation and drift detection to support alerting and root-cause analysis.
commercial · generally available · Research snapshot 2026-09-06
Show me drift and trust-and-safety metrics for a RAG application, including the retrieved source documents behind a flagged response.
Capabilities and evidence
Support labels reflect the supplied research. Documentation and vendor claims are not independent product tests. “Not established” means the researcher did not find support; it does not prove a capability is absent.
Documented by provider
Fiddler requires publication of LLM application inputs and outputs including prompts, prompt context, responses and, for RAG applications, retrieved source documents, and then generates trust and safety metrics and enrichments used for alerting, analysis and debugging.
Limit: The page does not document capture of LLM spans, tool calls or multi-agent handoffs, so agent-level traceability is unproven.
Documented capabilities include an embedding enrichment, drift detection and embedding visualisations with UMAP for root-cause analysis of problematic trends.
Limit: No documented retention, audit logging or access-control detail accompanying these analyses.
Product documentation describes monitoring traditional ML models, LLM applications and agentic applications in real time, experiments to test and validate LLM outputs, evaluation before and after deployment, and guardrails to protect AI applications.
Limit: Specific compliance certifications, RBAC design and policy-gating mechanics are not stated.