The autonomy paradox: why governing AI agents is what sets them free
Details
Abstract: Autonomous AI agents are no longer a research concept, they are scheduling meetings, writing code, and executing workflows right now. But autonomy without governance is a liability, not a feature. This session explores the spectrum of agent autonomy and what it takes to govern it responsibly. We'll look at where autonomous agents fail, what a layered governance stack actually looks like in practice, and why the organisations that invest in controls are the ones that end up moving fastest. Whether you're building agents or deciding whether to trust them, this session gives you the framework to do both with confidence.
Speaker: Manoj Jahgirdar is a Senior AI engineer with 7 years of experience in IT Industry building AI systems that run at enterprise scale. He is having expertise in AI architecture, multi-agent orchestration, Retrieval-Augmented Generation (RAG) pipelines, Large Language Model (LLM) fine tuning, evaluations, and monitoring AI systems. He writes about AI and technology on Medium and LinkedIn for both technical and business audiences, and mentor developers making the shift to AI-native engineering. He is also a strong believer of giving back to the developer community and he has contributed in various ways such as open-source contribution, hands-on labs for students, and much more. Learn more about Manoj work, patents, contributions or get in touch with him at https://manojjahgirdar.com.