Start the program with a clear overview of the learning journey, curriculum structure, program offerings, expectations, and key guidelines for success.

Build the Python foundation needed for AI development. Learn core programming, data structures, functions, OOP, file and error handling, AI-assisted coding, Pandas, NumPy, FastAPI, Pydantic, async Python, testing, and configuration

Build a clear foundation in GenAI and agentic AI. Understand how AI, ML, deep learning, transformers, LLMs, agents, tools, memory, and orchestration work together. Then explore frameworks such as LangChain, LangGraph, Deep Agents, and Microsoft Agent Framework.

Explore how modern LLMs have evolved from chatbots into multimodal, reasoning, and tool-using systems. Learn model selection, architecture, prompting, context management, chatbot development, and evaluation across text, code, math, audio, video, and multimodal benchmarks.

Learn how to build GenAI applications that go beyond text. This course helps learners work with LLMs and VLMs, use prompting, function calling, MCP tools, RAG, and RAGAS, then bring everything together by creating and deploying a multimodal image ad generator.

Learn to design AI systems that move beyond prototypes. Plan AI architectures, select appropriate workload patterns, design APIs and model gateways, and manage data, retrieval, caching, and infrastructure. Then optimize systems for latency, throughput, cost, reliability, and deployment in production environments.

Learn to move GenAI systems from prototype to production. Build reliable RAG pipelines, improve retrieval quality, serve models efficiently with caching and vLLM, control costs, monitor performance, govern MCP tools, and enable agent-to-agent communication.

Learn to design AI systems in which multiple agents work together. Compare RAG and agentic RAG, develop single-agent and multi-agent workflows, apply communication patterns, explore Google A2A, and orchestrate agents using LangGraph, CrewAI, and LangChain.

Learn to develop the runtime around an AI agent. Design complete agent harnesses across prompts, context, memory, tools, policies, skills, guardrails, logging, security, evaluation, and self-improvement loops using systems such as Hermes, Claude Code, LangGraph, and OpenAI Agents SDK.

Learn how advanced AI agents plan, act, observe, and improve through repeated loops. Design agent goals, develop swarm architectures, compare tools and frameworks such as Claude Code, Claude Agent SDK, Managed Agents, Qwen-Agent, and Kimi, and prepare agent systems with observability, permissions, cost controls, evaluation, and human approval.

Learn to build AI agents for enterprise environments. Design production-ready agent architectures, manage identity and access, defend against prompt injection and tool misuse, implement guardrails and evaluation, and deploy secure agents using Azure AI Foundry.

Bring everything together in a final, portfolio-ready project. Learners apply GenAI, agentic AI, system design, deployment, evaluation, security, and monitoring concepts to build a complete production-style AI solution.

Learn from Duke University Fuqua faculty through a live academic masterclass on advanced GenAI and agentic AI topics. Examine emerging trends, their business implications, and how AI is reshaping leadership and decision-making.