A podcast conversation between Shweta Vohra and Joseph Stein exploring the architectural and operational realities of building agentic AI systems in enterprises. Key topics include distinguishing truly agentic use cases from traditional automation, new security risks like prompt injection and tool hijacking, the need for
Table of contents
Watch the video:TranscriptNext Generation Architecture Playbook - Episode 3Agentic Systems Without Chaos [ 01:27 ]Why This Space Feels Different [ 02:27 ]Shift or Just More Automation? [ 03:36 ]What Counts as Truly Agentic [ 04:57 ]Open Source and the New Wave of Experimentation [ 08:21 ]What This Means for Architects and Designers [ 11:14 ]New Risks in Agentic Systems [ 16:12 ]Explainability Without Noise [ 20:39 ]More Power, More Work [ 24:47 ]What an Early Operating Model Looks Like ()[ 32:30 ]Merging the New with the Existing [ 40:02 ]Cost, Scale, and Sustainability [ 42:54 ]Hard Lessons from Early Success [ 48:19 ]What Comes Next [ 51:58 ]About the AuthorSort: