Skip to content
PlanAI

Welcome to PlanAI

A powerful Python framework for creating and orchestrating AI-enhanced workflows using type-safe, graph-based task automation
Get Started View on GitHub

Why PlanAI?

Section titled “Why PlanAI?”

Graph-Based Architecture

Build dynamic workflows with interconnected TaskWorkers for highly customizable automation

Type-Safe with Pydantic

Ensure data integrity and consistency across workflows with Pydantic-validated inputs and outputs

LLM Integration

Seamlessly combine traditional computations with LLM-powered operations and RAG capabilities

Built-in Monitoring

Track workflow execution in real-time with the integrated web dashboard and provenance tracking

Quick Example

Section titled “Quick Example”
from planai import Graph, TaskWorker, Task, LLMTaskWorker, llm_from_config


# Define a simple data processor
class DataProcessor(TaskWorker):
    output_types: List[Type[Task]] = [ProcessedData]


    def consume_work(self, task: RawData):
        processed = self.process(task.data)
        self.publish_work(ProcessedData(data=processed))


# Add AI analysis with an LLM
class AIAnalyzer(LLMTaskWorker):
    prompt = "Analyze the data and provide insights"
    output_types: List[Type[Task]] = [AnalysisResult]


# Create and run the workflow
graph = Graph(name="Analysis Pipeline")
processor = DataProcessor()
analyzer = AIAnalyzer(llm=llm_from_config("openai", "gpt-4"))


graph.add_workers(processor, analyzer)
graph.set_dependency(processor, analyzer)
graph.run(initial_tasks=[(processor, RawData(data="..."))])

Key Features

Section titled “Key Features”

Intelligent Routing

Type-aware data routing automatically manages flow between nodes

Provenance Tracking

Trace task lineage through the entire workflow for debugging

Caching Support

Built-in caching for expensive operations and LLM responses

Prompt Optimization

AI-driven automatic prompt improvement using production data

Next Steps

Section titled “Next Steps”

Ready to build your first AI workflow? Check out our guides:

Installation Guide

Get PlanAI installed and set up your development environment Get Started →

Quick Start Tutorial

Build your first workflow in minutes with our step-by-step guide Learn More →

Examples

Explore real-world examples including research assistants and Q&A generation View Examples →