phoenix

Arize Phoenix demonstrates using PXI to run an experiment comparing system prompt vs schema-aware prompt with a programmatic code evaluator, avoiding the need for an LLM judge.

Introduces how to use Phoenix's Pixie assistant to automatically classify error spans, batch annotate, and generate system prompt repair suggestions based on failure patterns.

Phoenix has a built-in Pixie assistant that helps users quickly filter out silent failure traces where agent spans have errors but model responses are normal, greatly improving trace reading efficiency.

Arize Phoenix announces trace-level annotations now display as peer columns in the trace header and as a section in the project stats panel, with live updates during streaming.

Phoenix 17.7.0 adds token detail charts that break down prompt and completion tokens into subcategories, with pan, zoom, and live-stream capabilities for better observability of AI model token usage.

Arize Phoenix's built-in agent PXI now supports slash commands and skills, allowing users to invoke custom workflows directly from chat.

A Waymo autonomous vehicle avoided a severe collision with a reckless wrong-way driver in Phoenix by taking evasive action, highlighting the safety capabilities of self-driving technology.

A developer reflects on building Migo Games, an online mini games app, using Elixir with Phoenix and Swift with SpriteKit, emphasizing the role of AI coding assistance and the scalability benefits of Elixir's process model.

Phoenix introduces Code Evaluators, allowing users to define evaluation strategies in Python or TypeScript directly in the UI, with server-side execution and composable scoring methods.