We publish cloud-native tech news, curate in-depth content, host regular events and live streams, and share Alibaba product and user case studies. Join us to explore and share the cloud-native insights you need.
A survey of 700+ developers reveals that while most enterprises have moved Agents into POC or production, critical gaps remain in observability, automated evaluation, optimization loops, and audit compliance, with strong demand for better reasoning quality and skill reuse.
UnifiedModel provides an open-source semantic layer that organizes enterprise assets, data, and relationships into a queryable object graph, enabling AI agents to read metrics by object and trace root causes along relationships; experiments on DataAgentBench show 10-20% accuracy gains for four flagship models, with GLM-5.2 reaching 50.2% pass@1.
This article analyzes DeepSeek Harness from a post-training perspective, revealing how its architecture bakes enterprise evolution into environment shaping and trajectory sedimentation, validated by a self-evolution POC that identifies interface design, contract shape, and feedback quality as critical bottlenecks.
The Shanghai finale of the Agent Observability and Optimization Close the Loop tour featured three technical sessions on building verifiable Agent optimization loops, managing Skills as strategic assets with SkillOps, and applying LangChain's Agent Engineering lifecycle from prototype to production, plus a hands-on workshop using Qoder and PawBench to demonstrate end-to-end evaluation.
The article argues that enterprises should invest in persistent business context rather than repeatedly rebuilding general Agent capabilities, using message-driven data integration and a unified semantic layer to make real-time, multi-source data reliably usable by Agents, illustrated by EventHouse's architecture.
This article details AgentLoop's experience self-evolution: automatically mining success/failure patterns from agent run traces, injecting them via a recall skill, and using ablation experiments to optimize recall threshold (0.6), top-1 injection, context positioning, experience slimming, and guardrails—achieving 30-40% latency reduction, 20-47% cost reduction, and significant token/tool-call savings.
This article shows how AgentLoop creates a backtesting plan, uses an offline intranet experiment platform, schedules regular runs, visualizes results on an experiment dashboard, and applies per‑question Rubrics to reliably measure and iterate on AI agent improvements.
This article walks through constructing a fully quantifiable and explainable evaluation system for AI agents—starting with business‑level golden metrics, using LLMs to break them into a detailed Rubric, embedding the Rubric in a custom evaluator, configuring evaluation tasks with trace data, and closing the loop by turning low‑scoring cases into actionable insights for continuous improvement.
This article explains how AgentLoop uses the OpenTelemetry protocol and probes to ingest high‑quality runtime data, offering four integration methods—one‑click generic agents, SDK framework integration, annotation‑based high‑code, and eBPF—demonstrated with a Claude Code customer‑service agent and end‑to‑end verification on the observation page.
The Shenzhen session of the Agent Observation & Optimization tour gathered nearly a hundred technologists to discuss evaluation paradigms, showcase AgentScope 2.0’s enterprise‑grade features, demonstrate a Java e‑commerce chatbot assessment with AgentLoop, and offer a hands‑on workshop, while previewing the upcoming Shanghai event.
