Netra

Phoenix is an open-source observability library designed for experimentation, evaluation, and troubleshooting. It allows AI engineers and data scientists to quickly visualize their data, evaluate performance, track down issues, and export data to improve. Phoenix is built by Arize AI, the company behind the industry-leading AI observability platform, and a set of core contributors. Phoenix works with OpenTelemetry and OpenInference instrumentation. The main Phoenix package is arize-phoenix. We offer several helper packages for specific use cases. Our semantic layer is to add LLM telemetry to OpenTelemetry. Automatically instrumenting popular packages. Phoenix's open-source library supports tracing for AI applications, via manual instrumentation or through integrations with LlamaIndex, Langchain, OpenAI, and others. LLM tracing records the paths taken by requests as they propagate through multiple steps or components of an LLM application.

Atla is the agent observability and evaluation platform that dives deeper to help you find and fix AI agent failures. It provides real‑time visibility into every thought, tool call, and interaction so you can trace each agent run, understand step‑level errors, and identify root causes of failures. Atla automatically surfaces recurring issues across thousands of traces, stops you from manually combing through logs, and delivers specific, actionable suggestions for improvement based on detected error patterns. You can experiment with models and prompts side by side to compare performance, implement recommended fixes, and measure how changes affect completion rates. Individual traces are summarized into clean, readable narratives for granular inspection, while aggregated patterns give you clarity on systemic problems rather than isolated bugs. Designed to integrate with tools you already use, OpenAI, LangChain, Autogen AI, Pydantic AI, and more.