Hamed Alikhani
Senior AI Engineer & Data Scientist at McGraw Hill
Hamed Alikhani, PhD is a Senior AI Engineer and Data Scientist working at the intersection of production AI engineering and applied statistical data science. He specializes in building scalable, real-world AI systems for high-stakes business problems, combining rigorous mathematical modeling with production-grade GenAI architectures. His work spans agentic AI systems, LLM safety and evaluation, retrieval-augmented generation (RAG), forecasting, optimization, and causal inference. Hamed has led impactful AI initiatives across industry and academia, delivering measurable business value through robust, explainable, and trustworthy AI solutions. He is also the Founder and President of Austin AI Hub, where he leads community-driven efforts to advance AI literacy, collaboration, and responsible AI adoption through workshops, panels, and hands-on programs.Watch in-person: February 18
From Prompts to Production: Building Reliable AI Agents for the EnterpriseAI agents are rapidly moving from experimental prototypes to critical components of enterprise systems. However, many organizations struggle to bridge the gap between prompt-based demos and reliable, production-ready agentic workflows. This talk explores how to design, build, and deploy AI agents that operate safely, efficiently, and at scale in real-world enterprise environments.
Drawing on hands-on experience deploying agentic AI systems across regulated and high-impact use cases, this session breaks down the core architectural patterns behind modern AI agents, including task decomposition, tool orchestration, memory management, and decision routing. Particular emphasis is placed on production challenges such as latency optimization, cost control, evaluation, observability, and LLM safety—areas that are often overlooked in early-stage implementations.
Attendees will learn practical lessons from real deployments, including how to move beyond brittle prompt chains, how to introduce guardrails and validation layers, and how to design agents that deliver measurable business value while remaining trustworthy and controllable. The session is designed for practitioners and leaders looking to move agentic AI from experimentation into dependable enterprise systems.
