dynamiq

Active
GitHub Python Apache-2.0

Description

Dynamiq is an orchestration framework for agentic AI and LLM applications.

Key Features

  • Orchestration framework for agentic AI and LLM applications with RAG support
  • Workflow engine supporting parallel and sequential multi-agent pipelines
  • Input transformers and node dependencies for complex data flow control
  • Async execution with first-class support for concurrent agent runs
  • Integration with E2B Code Interpreter and external tool sandboxes
  • Simple Python API with pip install and minimal boilerplate

Use Cases

πŸ’‘ Building multi-agent workflows with parallel and sequential orchestration
πŸ’‘ RAG pipeline construction for document Q&A and knowledge retrieval
πŸ’‘ Creating ReAct agents with tool-use capabilities for complex tasks
πŸ’‘ Enterprise AI application development with workflow automation
πŸ’‘ Prototyping and deploying LLM-powered applications with modular components

Quick Start

Install via `pip install dynamiq`. Create a simple LLM flow by importing OpenAI connection, defining a Prompt with Message objects, and instantiating an OpenAI node. Call `llm.run(input_data={...})` to execute. For agents, use `from dynamiq.nodes.agents import Agent` and wire tools via the `tools` parameter.

Related Projects