Building & Evaluating Agentic AI Systems

Learn the fundamentals of Generative AI, LLMs, and Agentic AI, along with prompting and RAG techniques. Explore real-world applications through hands-on demos across leading platforms like ChatGPT and Claude. Gain practical exposure to APIs, tools, and environment setup for building and deploying AI-powered systems.

1. 1. Introduction to Generative AI and Agentic AI

Learn prompt engineering fundamentals and advanced techniques like persona, few-shot, and chain-of-thought prompting. Explore multimodal capabilities across text, images, audio, and video with hands-on demos. Gain practical experience using LangChain to solve real-world problems, automate workflows, and build intelligent AI-driven applications.

1. 1. Prompt Engineering and LangChain Essentials

Learn the foundations of Retrieval-Augmented Generation (RAG) by building an end-to-end pipeline from scratch. This session covers document ingestion, text chunking, embeddings, vector databases, retrievers, and answer generation, with hands-on guidance to create a document-grounded QA system that produces accurate and context-aware responses.

1. 1. Build your first RAG Systems

Advance your RAG systems with techniques that improve retrieval quality and answer accuracy. Explore multi-query retrieval, keyword and hybrid search, reranking, contextual retrieval, and prompting strategies to reduce hallucinations. Apply these methods through practical projects to build robust, production-ready RAG applications.

1. 1. Build Advanced RAG Systems

Learn the fundamentals of Agentic AI by understanding what makes systems agentic and how LLMs, tools, prompts, tool-calling, and orchestration work together. Build your first ReAct AI agent using LangChain and LangGraph with built-in utilities, gaining hands-on experience with the agent reasoning and action loop.

1. 1. Build Your First ReAct AI Agent with LangChain & LangGraph

Build ReAct agents from scratch using LangGraph by working with states, nodes, edges, conditionals, and memory. Explore essential agentic design patterns such as routing, planning, reflection, and multi-agent workflows, then apply them to create scalable, real-world AI agents for complex problem-solving.

1. 1. Agentic AI Design Patterns & Building ReAct Agents from Scratch with LangGraph

Learn how to build conversational AI agents that remain reliable across long, multi-turn interactions through effective context engineering. Explore short-term and long-term memory, history management, memory trimming, tool output summarization, and context retrieval techniques while building a memory-aware Healthcare Assistant that delivers consistent, relevant responses.

1. 1. Memory Management & Conversational Agentic AI Systems

Discover how the Model Context Protocol (MCP) enables standardized integration between AI agents, tools, prompts, and data. Learn to build MCP servers and clients with FastMCP, understand multi-server MCP architectures, and explore how MCP supports scalable, enterprise-grade agentic systems across cross-functional workflows.

1. 1. Integrating MCP with Agentic AI Systems

Learn how to design production-ready multi-agent systems using LangGraph by exploring network, swarm, and supervisor-based orchestration patterns. Understand how agents collaborate, delegate tasks, and coordinate reliably using constructs like Command to solve complex workflows. Through hands-on examples, you'll build scalable, maintainable multi-agent applications for real-world enterprise use cases.

1. 1. Building Multi-Agent Systems with LangGraph

Discover how to build intelligent AI systems that combine retrieval, reasoning, and multimodal understanding. Learn to create Agentic RAG pipelines and multimodal agents capable of processing documents, PDFs, images, and other unstructured data. Apply these concepts in practical projects such as customer support, invoice processing, and enterprise knowledge navigation.

1. 1. Building Agentic RAG and Multimodal Agentic AI Systems

Learn how to systematically evaluate the performance of LLM and RAG applications using industry-standard metrics and frameworks. Build evaluation datasets, measure retrieval quality with contextual precision and recall, and assess response quality through answer relevancy, hallucination detection, and LLM-as-a-Judge techniques. Gain hands-on experience with tools such as Ragas, DeepEval, and Opik to create reliable evaluation workflows.

1. 1. Evaluating LLM & RAG Systems

Explore how to evaluate agentic AI systems by measuring goal completion, tool execution accuracy, and reasoning trajectories. Learn to add observability with tracing, monitoring, latency, and cost analysis using LangSmith to ensure production reliability. The session concludes by bringing together evaluation and monitoring best practices into a reusable workflow and reviewing the key concepts covered throughout the course.

1. 1. Building Agentic RAG and Multimodal Agentic AI Systems