> ## Documentation Index > Fetch the complete documentation index at: https://docs.budecosystem.com/llms.txt > Use this file to discover all available pages before exploring further. # Introduction to Observability > Monitor inference performance, investigate requests, and enforce protection rules ## Overview Bud AI Foundry Observability provides a single workspace for understanding inference behavior across models, deployments, projects, and users. It combines metrics, request-level diagnostics, and enforcement rules so teams can monitor service health and respond quickly. ```mermaid theme={null} flowchart LR A[Inference Traffic] --> B[Metrics Tab] A --> C[Requests Tab] A --> D[Rules Tab] B --> E[Performance Insights] C --> F[Request Investigation] D --> G[Traffic Protection] ``` ## Key Features **Inference-Centric Metrics** Track requests, latency, token usage, TTFT, and throughput over configurable time windows. **Request-Level Debugging** Drill into individual requests with metadata, payload details, and trace-ready identifiers. **Rules & Blocking Controls** Manage IP, country, user-agent, and rate-based rules with active/inactive state and lifecycle actions. **Consistent Filtering** Apply time range, text search, and scoped filters to narrow investigations quickly. Image ## Core Workflows 1. **Monitor** trends on the Metrics tab. 2. **Investigate** anomalies from the Requests tab. 3. **Protect** traffic using Rules tab controls. 4. **Iterate** by adjusting filters and repeating analysis. ## Getting Started Explore metrics and requests in minutes Learn key objects, signals, and tabs Run a practical investigation flow