LangChainAI has announced the release of LangGraph.js 0.2, a controllable orchestration framework designed to build reliable AI agents in JavaScript and TypeScript. The new version includes features such as dynamic breakpoints, parallelization, and the ability to flexibly stream intermediate steps, rewind and debug past states, and add human-in-the-loop to approve actions. LangGraph Studio and Cloud have also been introduced, enabling local development and seamless production deployment. A companion YouTube video and blog post provide step-by-step guides and examples for using the new tools.
Build and deploy AI agents in Javascript using LangGraph js and visualize them in LangGraph Studio (AI agent IDE) https://t.co/pVQjDEr70b
Really useful open-sourced tool for evaluating and enhancing LLM deployments by @neuralmagic - GuideLLM ✨ 🔍 Simulates real-world inference workloads 🏋️ Performance evaluation under various loads 💻 Resource optimization for hardware configs 💰 Cost estimation for deployment… https://t.co/j6n6vTX58h https://t.co/NYHixU442K
🙋 Day 2 of LangGraph.js week: Human-in-the-loop with JavaScript agents 🙋 Letting an LLM decide what code to run can unlock incredibly powerful user experiences, but for sensitive tasks, you may want human approval as a safeguard. LangGraph.js allows you to set both fixed and… https://t.co/GvVYKCkgho