Custom AI Workflow

AI Learning Guide

Aralia OpenRAG SDK

The Aralia OpenRAG SDK is publicly available on GitHub. You can clone the repository with any Git tool (e.g., SourceTree) to read the code and practice.

https://github.com/araliadata/AraliaOpenRAG

Go further: Create (or modify) your own workflow

If you’d like to build your own AI workflow, start with the references below and tailor an AI workflow that best fits your use case.

Where can I start making changes?

Flow Orchestrator: graph.py — uses LangGraph to chain the five nodes and run them in a fixed order.
Node Logic: node.py — core behaviors of the five agents.
Tools Layer: aralia_tools.py — OAuth, dataset search, field/range exploration, and actual querying.
Data Schema: schema.py — fields, formats, and calculation methods.
Prompt Templates: prompts.py — dataset selection, chart planning, and insight interpretation.
State Management: state.py — LangGraph state, including question, ai, at, final_response, etc.

New to Python? Where to learn

Official Tutorial: https://docs.python.org/3/tutorial/
Google Python Class (free): search for “Google Python Class”
Online courses: Coursera (Python for Everybody), edX, Codecademy, etc.

<aside> 📌

Tip: Get comfortable with: variables, functions, conditionals, loops, modules, installing packages with pip, and running notebooks in Colab.

</aside>

← Previous: Experiment with Aralia SDK in Colab

Next: 🔌 DataPlanet API →