Ready to become a certified watsonx Data Scientist? Register now and use code IBMTechYT20 for 20% off of your exam →
Learn more about agentic workflows with NirvanAi here →
🚀 How do AI agents transform workflow orchestration? Terzo President and COO Eric Pritchett explains how LLMs and AI agents simplify automation, boost efficiency, and enable innovation. Discover how orchestration is redefining traditional RPA and driving smarter processes.
AI news moves fast. Sign up for a monthly newsletter for AI updates from IBM →
#aiagents #llm #rpa #aiautomation
The RPA vs orchestration comparison really crystallizes the key difference – RPA needs every step explicitly defined, while orchestration with LLMs can handle the ambiguity in between steps. One thing I’ve been exploring in production is adding per-step cost budgets to the orchestration layer so each agent in the workflow has a token ceiling. Prevents a single reasoning-heavy agent from consuming the entire inference budget. Has anyone experimented with tiered model routing within their orchestration pipelines – like using a smaller model for validation steps and reserving the larger model for the actual reasoning?
Anyone know how the whiteboard writing in the video is done? Thanks
After watching this… I’ll just write the script..
Shouldn’t the MCP for the CRM, Product SKUs etc. applications be a Server rather than a Host? I think the host is the application rather than the “services” it consumes.
That’s beautifully explained! 🙏🏼
Excited to see how it would in real life situations.