Also available on:
In this episode of GydeBites, host Prasanna Vaidya speaks with Jigar Shah on why building agentic AI systems requires far more than powerful models and impressive demos. Jigar explains how reliability, governance, auditability, and evaluation frameworks separate production-ready systems from experiments.
The conversation covers multi-agent orchestration, trust, human oversight, context management, and the engineering practices needed to deploy AI systems safely at scale.
This episode offers practical guidance for teams looking to build agentic AI systems that are dependable, governed, and enterprise-ready
Where is noise, and where is rubber hitting the road in agentic systems?
Why do so many agentic AI projects fail?
What data and governance gaps make agentic systems unreliable in enterprises?
What should enterprises do to overcome agentic AI challenges?
What are some of the traits of a production-grade agentic system vs an AI experiment?
Head of Applied AI & Platform Engineering, Atomicwork
With over 20 years of experience across AI, enterprise software, and data platforms, with leadership stints at Amazon's AGI organization and ServiceNow, and business acumen forged at McKinsey & Company and Chicago Booth. His current focus is multi-agent orchestration, autonomous AI workforce, and production-grade AI that earns trust through performance, not hype.
