Designing Agentic Systems with LangChain
| Course Details | ||||||||
|---|---|---|---|---|---|---|---|---|
|
||||||||
| Languages | ||||||||
|
||||||||
| Outcomes | ||||||||
|
||||||||
| Price & Terms | ||||||||
|
||||||||
| Technical | ||||||||
|
||||||||
| Specification | ||||||||
|
||||||||
Build ReAct agents and LangGraph chatbots in Python that call custom tools, query Wikipedia and keep memory across multi-turn conversations.
Agentic systems pair a language model with tools so it can reason about a task, call the right function and act on the result. You set up ReAct agents in Python with LangChain and give them tools of your own. Then you build chatbots as LangGraph graphs that retrieve real-time information through external APIs, keep memory across turns and choose among several tools as a conversation unfolds. The aim is a chatbot that moves smoothly between plain replies and tool-assisted answers.
What You Will Achieve
- Configure a ReAct agent that reasons over a question and calls a Python function you wrote to handle math calculations.
- Lay out a chatbot as a LangGraph graph and generate its diagram to check how the conversation flows.
- Connect a Wikipedia lookup to a chatbot so it answers with retrieved information when a question calls for it.
- Give one chatbot several tools, route each request with conditional edges and keep memory so a multi-turn chat stays coherent.
Program Overview
Essentials of LangChain Agents
How prompts, LLMs and tools combine in a ReAct agent, and how to give that agent tools of its own.
- ReAct agent setup with OpenAI’s API
- Custom tools, including a math calculator
- Agent conversations and questions about history
- Graphs, nodes and edges in LangChain
Chatbots with LangGraph
A chatbot built as a graph whose states and edges let it switch between conversation and tool-assisted responses.
- Graph and agent states
- Response generation and graph diagrams
- A Wikipedia tool added to the graph
- Graph memory for conversation
Dynamic Chat Agents with Multiple Tools
Several tools bound to one model, with functions that decide which tool runs and when the chatbot stops.
- Tools that invoke the LLM or run Python code
- Binding multiple tools
- Nodes and edges for flexible function calling
- Organized outputs and conversation memory
Certificate Awarded (Paid)
Job Outlook
Agent design with LangChain and LangGraph is used in AI engineering, data science and data engineering work that automates multi-step tasks, builds chat assistants or links language models to external tools and APIs.
Good for: Python developers who want chatbots that call tools, keep memory across turns and follow workflows they lay out as LangGraph graphs.
Keep in mind: The exercises use small tools, such as a math calculator and a Wikipedia lookup, so applying these patterns to a business system is left to you.
Bottom line: A hands-on path through ReAct agents and LangGraph chatbots that ends with one bot handling several tools and remembering the conversation.
Frequently Asked Questions
What is an agentic system in this course?
Here it means a language model linked to tools and a workflow, so it can reason about a request, decide which tool to call and act on what comes back. The ReAct agent in the first chapter is the starting example.
Which tools do the chatbots use?
The exercises build a calculator tool for math, a Wikipedia tool for retrieving information, a tool that calls the LLM itself and one written in plain Python code. The final chatbot binds several of them and chooses among them as the conversation goes.
How does LangGraph fit alongside LangChain?
LangChain supplies the agent pieces, meaning prompts, the language model and tools, while LangGraph arranges a chatbot as a graph of states, nodes and edges. The lessons also generate a diagram of that graph so you can see how replies and tool calls connect.
Which language model service do the exercises use?
The agent in the first chapter is set up with OpenAI's API, and the later chatbots call the language model to generate their responses and inside one of their own tools.
User Reviews
Write your review
You must be logged in to post a review.
Reviews can only be posted by registered learners. Log in or create a free account, then you will return here to post.
There are no reviews yet.