Bind AI.
Build Anything.

The open-source Python AI framework

Build AI applications with agents, workflows, tools, memory, knowledge, providers, and project-based architecture — all designed to work together in Python.

BindAI is an open-source Python framework for building AI applications that go beyond simple model calls. It provides a modular architecture for creating agents, orchestrating workflows, connecting tools, managing memory and knowledge, integrating model providers, and organizing applications into reusable projects.

Whether you're building an AI assistant, document processing pipeline, business automation system, knowledge-powered application, or multi-agent solution, BindAI provides the core components needed to develop and extend real-world AI software.

Why BindAI?

Modern AI applications require more than a language model. BindAI brings together the core building blocks needed to create modular AI software:

Agents — Build AI agents that process requests, use tools, and produce application-level results.

Tools — Extend agents with functions, APIs, external services, and application-specific business logic.

Memory — Provide applications with persistent and conversational context for stateful AI experiences.

Knowledge & RAG — Connect applications to documents, embeddings, retrieval systems, and external knowledge.

Workflows — Orchestrate execution with sequential nodes, conditions, loops, parallel branches, retries, timeouts, human tasks, and scheduling.

Projects — Organize applications, agents, workflows, configuration, and shared resources into structured AI projects.

Core Principles

BindAI is built around a few foundational principles:

Python-first — Build AI applications using clean, idiomatic Python.

Modular — Use framework components independently or compose them into larger applications.

Provider-agnostic — Integrate different AI model providers without coupling application logic to a single provider.

Workflow-driven — Coordinate AI execution through explicit, inspectable workflow nodes and execution patterns.

Developer-friendly — Keep application logic readable with clear APIs, reusable components, and minimal unnecessary abstraction.

Architecture

A typical BindAI application is structured as follows:

Project
│
├── Applications
│ ├── Agents
│ ├── Tools
│ ├── Memory
│ └── Knowledge
│
├── Workflows
│
├── Shared Tools
│
└── Scheduler

Each component has a single, clear responsibility, making the system easier to reason about, extend, and maintain over time.

Documentation

The BindAI documentation is structured to guide developers from the core concepts to advanced workflow and application patterns:

• Getting Started — Install BindAI and build your first AI application.
• Core — Understand the main framework concepts and architecture.
• Tools — Extend agents and applications with custom capabilities.
• Memory — Build stateful applications with reusable context and memory components.
• Knowledge — Connect AI applications to knowledge and retrieval systems.
• Workflows — Orchestrate sequential, conditional, looping, parallel, retry, timeout, human, and scheduled execution.
• Projects — Organize AI applications and their supporting components.
• Templates — Learn framework patterns through complete workflow and application examples.
• API Reference — Explore the BindAI package APIs.

Quick Example

from bindai import Agent

agent = Agent.from_yaml("assistant.yaml")

result = agent.run(
"Explain quantum computing in simple terms."
)

print(result.output)

As your application evolves, you can incrementally add tools, workflows, memory, knowledge retrieval, scheduling, and human approval without changing your core architecture.

Who Is BindAI For?

BindAI is built for:

• Python developers

• AI engineers

• AI application developers

• Business automation teams

• Internal AI application builders

• Workflow automation developers

• Knowledge and retrieval application builders

• Multi-agent system developers

If you're building AI software with real application logic, orchestration, and reusable components — not just experimenting with prompts — BindAI is designed for you.

Next Steps

New to BindAI? Start here:

Read the Getting Started guide
Build your first Agent
Create your first Workflow
Explore the Templates
Explore the API Reference

From there, you can gradually introduce tools, memory, knowledge retrieval, conditions, loops, parallel execution, retries, timeouts, human tasks, and scheduling as your applications grow in complexity.

The open-source Python framework for AI applications.

Stay in the loop

Bind AI.Build Anything.