Unlock the New Flow Builder in IBM watsonx Orchestrate – Visualize, Debug & Deploy Tool Flows скачать в хорошем качестве

Unlock the New Flow Builder in IBM watsonx Orchestrate – Visualize, Debug & Deploy Tool Flows

Ready to build smarter AI Agents faster? 🔧 This video walks you through the brand-new Flow Editor in the watsonx Orchestrate Developer Edition—a visual, code-enhanced way to design, test, and debug agent tool flows locally. Perfect for developers working with the Agent Development Kit (ADK) and Langfuse! WARNING Custom Docker-Compose in watsonx Orchestrate it not official supported: https://developer.watson-orchestrate.... Before starting watsonx Orchestrate Developer Edition with a custom Docker Compose file, make sure you understand every change in your configuration. The ADK doesn’t offers official support to custom Compose setups, so you’re responsible for troubleshooting any issues that arise. 🎯 Whether you're just getting started or already deep into building with agents, this short hands-on video is your step-by-step launchpad—from setup to debugging. 👁️‍🗨️ Visually design tools. 🧠 Add code logic with ease. 🐞 Catch bugs using Langfuse. ⚡ Test with different agent architectures (default & ReAct). Table of content : 00:00 – Intro & Setup Recap : Kick off by referencing the blog post “Getting Started with Local AI Agents in the watsonx Orchestrate Developer Edition” which walks through installing the ADK, Docker setup, local server start, and optional Langfuse observability. 01:43 – Navigate to the Flow Editor : Show where to access the visual Flow (Agent) Editor in the local UI. 02:51 – Create the SimpleCounter Agent : Demonstrate initiating a new agent named “SimpleCounter.” 03:53 – Build a Flow Tool Called “SimpleCounter” : Add a flow based tool, enabling drag-and-drop flow design. 04:37 – Define the Flow Input : Add an input that accepts a numeric counter value from the user. 05:38 – Add a Code Block : Insert a code block where the counter logic will reside. 06:17 – Add Output to Code Block : Define the output interface for the code logic. 06:55 – Implement Code in Block : Write the increment operation inside the code . 07:20 – Add Flow Output Node : Attach an output node to return the incremented value. 08:10 – Run Agent with the default prompt Architecture : Test-execute the flow using the default . 10:02 – Run Agent in ReAct Architecture : Switch and run the flow using React Agent Architecture. 10:45 – Inspect an Error in Langfuse : Inspect the trace via Langfuse observability. 13:12 – *Error Identified * : Visual pinpoint of failure in the flow or code logic based on trace output. 13:32 – Summary & Outro : Recap core takeaways: setup → build → run → debug—all visually with Flow Editor and Langfuse. Related resources of suedbroecker.net “Getting Started with Local AI Agents in the watsonx Orchestrate Developer Edition” https://suedbroecker.net/2025/06/25/g... "Supercharge Your Support: Example Build & Orchestrate AI Agents with watsonx.ai and watsonx Orchestrate" https://suedbroecker.net/2025/05/15/s... "watsonx Orchestrate" https://www.ibm.com/docs/en/watsonx/w... #watsonx #AIagents #Langfuse #DeveloperTools #IBMAI #FlowEditor #OpenSourceAI #AgentDevelopment #ThomasSuedbroecker #watsonxOrchestrate #ReactAgentArchitecture #ADK