Parallel tool execution in Agentic AI

Introduction

Parallel tool execution helps run many tasks at the same time. This saves time and makes work faster.

When you want to ask multiple tools questions at once.

When you need to get answers from different sources quickly.

When you want to speed up a process by doing many steps together.

When you have many small jobs that can run without waiting for each other.

When you want to improve efficiency in AI workflows by multitasking.

Syntax

Agentic AI

results = agent_executor.invoke_parallel([tool1, tool2, tool3], inputs_list)

agent_executor.invoke_parallel runs tools at the same time.

inputs_list is a list of inputs matching each tool.

Examples

This runs a search and a calculator tool at the same time with different inputs.

Agentic AI

results = agent_executor.invoke_parallel([search_tool, calculator], ["weather today", "2+2"])

This runs translation and summarization tools in parallel.

Agentic AI

results = agent_executor.invoke_parallel([translator, summarizer], ["Hello", "Long article text"])

Sample Model

This code defines two simple tools as functions. It runs them at the same time using threads. Then it prints each tool's result.

Agentic AI

from concurrent.futures import ThreadPoolExecutor

# Define two simple tools as functions
def tool1(input_text):
    return f"Tool1 processed: {input_text}"

def tool2(input_text):
    return f"Tool2 processed: {input_text}"

# Function to run tools in parallel
def run_tools_parallel(tools, inputs):
    with ThreadPoolExecutor() as executor:
        futures = [executor.submit(tool, inp) for tool, inp in zip(tools, inputs)]
        results = [future.result() for future in futures]
    return results

# Tools list and inputs
tools = [tool1, tool2]
inputs = ["data A", "data B"]

# Run tools in parallel
outputs = run_tools_parallel(tools, inputs)

# Print results
for output in outputs:
    print(output)

Important Notes

Parallel execution can speed up tasks but needs careful input-output matching.

Not all tools can run safely in parallel if they share resources.

Summary

Parallel tool execution runs many tools at once to save time.

Use it when tasks do not depend on each other.

It improves efficiency in AI workflows by multitasking.