​

Understanding Process Types

Process types in PraisonAI define how tasks are executed and how agents collaborate. Each process type offers different patterns for task execution and agent coordination.

​

Process Types Overview

Three types of processes are implemented in PraisonAI:

​

Sequential Process

The simplest form of task execution where tasks are performed one after another.

agents = PraisonAIAgents(
    agents=[agent1, agent2],
    tasks=[task1, task2, task3],
    process="sequential"
)

​

Hierarchical Process

Uses a manager agent to coordinate task execution and agent assignments.

agents = PraisonAIAgents(
    agents=[manager_agent, worker_agent1, worker_agent2],
    tasks=[task1, task2],
    process="hierarchical",
    manager_llm="gpt-4o"
)

​

Workflow Process

Advanced process type supporting complex task relationships and conditional execution.

task_workflow = {
    "start_task": {
        "next": ["decision_task"],
        "type": "task"
    },
    "decision_task": {
        "type": "decision",
        "conditions": {
            "success": ["process_task"],
            "failure": ["error_task"]
        }
    }
}

​

Advanced Features

​

State Management

• Task state tracking
• Context preservation
• Data flow control

​

Error Handling

• Graceful failure recovery
• Alternative path execution
• Error reporting

​

Monitoring

• Progress tracking
• Performance metrics
• Resource utilization

​

Integration

• External system coordination
• API synchronization
• Event handling

​

Async Process Execution

All process types (sequential, workflow, hierarchical) support async execution through async generators. This enables efficient handling of async tasks and parallel processing.

​

Async Process Methods

• asequential: Async version of sequential process
• aworkflow: Async version of workflow process
• ahierarchical: Async version of hierarchical process

​

Example Usage:

import asyncio
from praisonaiagents import Agent, Task, PraisonAIAgents

async def main():
    # Create agents and tasks
    agents = [async_agent1, async_agent2]
    tasks = [async_task1, async_task2]

    # Initialize PraisonAIAgents with async workflow
    agents = PraisonAIAgents(
        agents=agents,
        tasks=tasks,
        process="workflow"  # or "sequential" or "hierarchical"
    )

    # Start async execution
    result = await agents.astart()
    print(result)

# Run the async workflow
asyncio.run(main())

​

Async Process Features:

• Parallel task execution in workflow
• Async task delegation in hierarchical process
• Non-blocking task execution
• Efficient resource utilization
• Mixed sync/async task support

​

Process-Specific Async Behavior:

1. Sequential Process:

   • Tasks are executed in order
   • Async tasks don’t block other tasks
   • Maintains sequential order while allowing async execution

2. Workflow Process:

   • Parallel execution of independent tasks
   • Async branching and merging
   • Dynamic task scheduling

3. Hierarchical Process:

   • Async manager delegation
   • Parallel subtask execution
   • Dynamic task assignment