metadata

language: code
license: mit
tags:
  - machine-learning
  - ai
  - structured-planning
  - llamaindex
model_name: Structured Planning AI Agent
model_type: agent
library_name: llama-index

Implementing a Structured Planning AI Agent with LlamaIndex

Set up the environment
- Skip this step if you have already set up the environment
```
python -m venv .venv
source .venv/bin/activate
```
Setup LlamaIndex
```
pip install llama-index
```
Create a python file
```
touch worker.py
```
Or
```
echo. > worker.py
```
Open the file in VSCode
```
code worker.py
```

Add the needed imports

from llama_index.core.tools import FunctionTool
from llama_index.llms.openai import OpenAI
from llama_index.core.agent import (
    StructuredPlannerAgent,
    FunctionCallingAgentWorker,
)

Define the function

def multiply(a: int, b: int) -> int:
 """Multiply two integers and returns the result integer"""
 return a * b

Define and configure the worker agent

multiply_tool = FunctionTool.from_defaults(fn=multiply)
llm = OpenAI(model="gpt-4o-mini")
worker = FunctionCallingAgentWorker.from_tools([multiply_tool], llm=llm, verbose=True)
worker_agent = StructuredPlannerAgent(worker, [multiply_tool], verbose=True)

Test the worker agent

worker_agent.chat("Solve the equation x = 123 * (x + 2y + 3)")

Create .env file & add api key
```
OPENAI_API_KEY="<your_api_key>"
```
Run the agent

python worker.py