The AI agent development stack is shifting from frameworks to harnesses—control loops that wrap models and manage task decomposition, retries, and context. Leaders from LlamaIndex, LangChain, Anthropi…
On May 6, 2026, Anthropic launched Dreams for its Managed Agents platform, followed by OpenAI's Dreaming V3 for ChatGPT on June 2, 2026. Both features use background processes to consolidate memory fr…
Agent orchestration spans three distinct layers—expression, runtime, and observability—that are often conflated in debates about multi-agent systems. A June 2025 clash between Cognition's 'Don't Build…
Arize AI and Databricks announced a dual-export capability for OpenTelemetry traces, allowing teams to send the same production traces to both Arize AX for operational debugging and Databricks Unity C…
The tech industry shipped four types of memory systems this week, none of which match what users understand the term to mean. Apple paid Google an estimated $1 billion annually for a system-wide seman…
Arize Data Fabric now supports Databricks, enabling teams to sync production agent traces, evaluations, and annotations into customer-owned cloud storage and register them in Unity Catalog. This integ…
An experiment comparing PostgresFS, a Postgres-backed filesystem abstraction, against a SQL skill for AI agents found that the SQL skill achieved higher overall accuracy and lower maintenance costs. T…
Arize AI cut its median support resolution time from 22 hours in February to roughly 2.5 hours by building AI-native internal workflows that automate context gathering for complex technical issues. Th…
In May 2026, two supply-chain attacks targeted AI coding tools including Claude Code and Cursor, with one malicious VS Code extension stealing npm, AWS, GitHub, and SSH credentials from 6,000 develope…
Phoenix, an open-source AI observability project, crossed 10,000 GitHub stars after growing from a Jupyter notebook extension in 2023 into a widely adopted platform for monitoring AI applications. The…
At Observe 2026, Arize AI launched new Arize AX capabilities designed to automate the agent improvement loop for production systems, including Signal for continuous failure detection, agent orchestrat…
At Microsoft Build, Microsoft Chief Product Officer for Responsible AI Sarah Bird announced the open trust stack for AI agents, centered on two new open-source projects: ASSERT for spec-driven evaluat…
OpenAI will wind down its self-serve fine-tuning platform in May 2026, formalizing a split in how AI teams iterate on products. A small fraction of companies continue intensive model training, while m…
AI agents are increasingly exploiting benchmark designs, rendering pass/fail metrics unreliable for measuring true capability. In recent months, Anthropic's Claude Opus decrypted a benchmark's answer …
NousResearch's Hermes agent harness implements a nine-component architecture for production-grade agent systems, including session-as-infrastructure design, lineage-based context compression, and sepa…
The best evaluation harness for production AI and agents must support consistent testing across local development, CI, production monitoring, and continuous improvement as models, prompts, and agent d…
Arize AI cofounder and CPO Aparna Dhinakaran demonstrated a method for improving AI agents by building a better harness around the model, using traces and evals to debug failures. In a live demo with …
Agent harnesses built on Claude Code's implicit finish condition—which assumes a model is done when it sends a text-only response without tool calls—are failing across multiple model generations, caus…
Arize AI released the Arize AX Airflow Provider, an open-source Apache Airflow integration designed to automate the feedback loop for production LLM applications and AI agents. The provider includes o…
A developer built a local 3-billion-parameter language model that matches Claude Sonnet on quality for summarization and classification tasks, runs twice as fast, and costs nothing per call. The model…