Gen-AI Developer Classroom notes 22/Oct/2025

Flight Checkin & Upgrade

• Overview
  

• To solve most of the business or enterprise needs we need workflows

Langgraph

• Langgraph is designed around graphs which make it feasible to run business workflows
• Main Components of Langgraph (w.r.t) Development
  • Node: A node is a python function or runnable.
    
  • Edge: This is directional flow
    
  • State: This is data which is shared to all the nodes

Setting up a project with langgraph

• Create a new folder called as hello-graph and cd into it
• Lets create a python project and dependencies

uv init .
uv add langgraph langgraph-cli[inmem]
uv sync

• Activate virtual environment

# windows
.venv/Scripts/activate

# linux or mac
source .venv/bin/activate

• Now open vscode

code .

• Lets create a simple workflow
  
• change the main.py code to the following

from langgraph.graph import START, END, StateGraph
from typing import TypedDict


class MyState(TypedDict):
    name: str
    message: str


def say_hello(state: MyState) -> MyState:
    """This is wish node

    Every node in langraph takes state as input and returns 
    state

    Args:
        state (MyState): State

    Returns:
        MyState: State
    """

    state['message'] = f"hello {state['name']}"
    return state

# design your graph
state_graph = StateGraph(MyState)
state_graph.add_node("wish", say_hello)
state_graph.add_edge(START, "wish")
state_graph.add_edge("wish", END)

# compile your graph
graph = state_graph.compile()

• Now create a langraph.json file with following content

{
    "dependencies": ".",
    "graphs": {
        "main": "main:graph"
    }

}

• Now in the activate virtual environment execute the following command

langgraph dev --allow-blocking

• It would open a browser if not http://localhost:2024
  

• Refer Here for the hello-world langgraph

Lets build a graph which does add two numbers

• Lets create a new file the same project above called as calc.py