This course provides a comprehensive overview of building intelligent agentic applications using LangChain. Participants will learn to create simple LangChain applications and explore structured outputs.
The curriculum progresses through multi-step workflows, transitioning from LLM calls to developing agents, and covers essential agentic design patterns. Key concepts include extending agents with tools, managing functions, and understanding state management within LangGraph.
Students will also implement short-term agent memory techniques and complete a hands-on project:
building a report-generating agent.
- Course Introduction & Setup
Get introduced to the course, meet your instructors, and learn how to set up and use your Vocareum OpenAI API key and assigned budget for hands-on activities.
- Creating a Simple LangChain application
Learn to build a simple LangChain app: integrate LLMs, manage chat history, use prompt templates, and apply few-shot prompting to create customizable chatbots with memory.
- Structured Outputs
Discover structured outputs in AI: transform responses into actionable JSON for integration. Utilize schemas, parsers, and function calls to enhance reliability and automation in workflows.
- LangChain Structured Outputs
Learn to parse and structure LLM outputs in LangChain using output parsers, TypedDict, Pydantic models, and automatic error correction for robust workflows.
- Multi-Step Workflows in LangChain
Learn to build flexible, multi-step AI workflows in LangChain using Runnables and LCEL for composing, batching, and managing complex chains with ease and scalability.
- From LLM Calls to Agents
Explore how AI evolves from basic LLM calls to fully autonomous agents, learn agentic frameworks, applications, and how to design effective agent-driven workflows.
- Agentic Design Patterns
Discover four key agentic design patterns—reflection, tool use, planning, and multi-agent collaboration—to boost LLM performance and reliability in complex tasks.
- Extending Agents with Tools
Extend AI agents beyond text with tool integrations, enabling reliable real-time actions and data access.
- Functions as Tools in LangChain
Learn how to encapsulate Python functions as tools in LangChain, enabling LLMs to call functions, interact with APIs, and handle external computations in AI workflows.
- Agentic Workflows with LangGraph
Learn to build dynamic, agent-driven AI workflows using LangGraph, leveraging nodes, edges, and routing for modular, adaptive application control and automation.
- Agent State Management
Explore agent state management with state machines. Learn how agents track user input, instructions, and tool use for complex workflows, ensuring adaptability and reliability.
- LangGraph State Management
Learn structured state management in LangGraph for robust AI workflows, with focus on TypedDicts, reducers, routing, handlers, persistence, and practical multi-turn agent design.
- Short-Term Agent Memory
Explore short-term memory in AI agents, enhancing coherence via state, ephemeral, and ephemeral memory strategies for efficient context retention in active sessions.
- LangGraph Agent Memory
Learn how LangGraph uses checkpointers and threads for agent memory, allowing persistent, isolated, and customizable conversations with snapshots, time travel, and personalized assistance.
- Project: Report-Building Agent
Build a document processing system using LangChain/LangGraph. You'll create an AI assistant that answers questions, summarize documents, and perform calculations on financial and healthcare documents.
