Advanced Bobchestrate Workshop¶

Advanced Topics in watsonx Orchestrate Agent Development¶

Welcome to the advanced workshop. These modules assume you have a working watsonx Orchestrate SaaS environment and the ADK CLI configured.

Before starting any advanced part:

✅ wxO SaaS environment active and orchestrate env list shows your environment
✅ ADK CLI installed and up to date: pip install --upgrade ibm-watsonx-orchestrate
✅ Python 3.11–3.13 with uv installed
✅ IBM Bob IDE installed and open
✅ IBM Bob configured to work with watsonx Orchestrate ADK
✅ Comfortable with Python tools, agents, and the wxO ADK CLI

Part	Topic	Time	Difficulty
0 — Setup & Environment	Get your environment ready: Bob IDE, MCP servers, wxO connection, API keys	15–20 min	⭐
1 — LangGraph Agents	Build, package, and deploy custom LangGraph agents with memory and tool-calling	60–75 min	⭐⭐⭐
2 — Event-Driven AI Agents	Confluent Cloud + Flink SQL + watsonx Orchestrate agent — real-time inventory intelligence	75–90 min	⭐⭐⭐⭐
3 — Scheduling (coming soon)	Make agents and workflows run on a schedule via natural language	20 min	⭐
4 — Advanced Workflow Nodes (coming soon)	Human-in-the-loop, Prompt nodes, Parallel execution, Doc Processing	30 min	⭐⭐
5 — LLM Model Policies (coming soon)	Fallback, load-balancing, and retry across multiple LLM providers	25 min	⭐⭐
6 — Langflow Tools (coming soon)	Build visual AI pipelines and import them as wxO tools	30 min	⭐⭐
7 — CI/CD & GitOps (coming soon)	Deploy agents through GitHub Actions using a GitOps pipeline	40 min	⭐⭐

Throughout the advanced modules, ask Bob for help:

"Bob, I'm building a LangGraph agent for wxO.
 Help me add a tool that calls the GitHub API using a wxO connection."

Bob knows the wxO ADK rules and the platform limitations. Ask it to explain concepts, generate code, debug errors, or suggest improvements.