LLM Applications — TraceVox

Route requests through the TraceVox gateway and get complete traces, live triage of incidents, cost and latency attribution, safety evaluations, and policy guardrails.

• Gateway: drop-in OpenAI-compatible proxy, authenticated by project API key.
• Distributed tracing: model events, latency breakdown, token and dollar cost.
• Live requests, model health, cost & risk, alerts, custom dashboards.
• Evaluations against curated safety datasets; datasets & regression testing; prompt playground.
• Triage queue (Early Warning Engine) and SAFE-mode guardrails; PII-safe tracing and audit logs.

LLM Applications docs · Voice Agents · Sign in

TraceVox Research home · Public Research Library · Documentation · llms.txt