Use Workflows in Python

Discover how to get started and use the workflows in Python

🛠️ How to Build Workflows

TuringDB lets you design and execute analytical and agentic workflows that combine:

• File processing
• LLM queries
• Graph analytics
• Graph memory via TuringDB

This guide shows how to use these workflows step by step using the workflowcommon Python SDK.

📦 Installation

To get started, clone the workflowcommon repository and install it using uv:

git clone <https://github.com/turingdb/workflowcommon.git>
cd workflowcommon
uv pip install -e .

⚙️ Workflow Usage Basics

We use the FlowBuilder class to define workflows using nodes and connections. Each node corresponds to a task (e.g., extract text, run an LLM, query TuringDB).

🔄 Steps to use a workflow:

1. Import nodes and builder
2. Create the workflow builder
3. Define and add nodes
4. Connect the nodes
5. Build and visualize the pipeline
6. Execute the pipeline
7. Get outputs and results