AI Agent Tutorial

AI Learning Guide

Summary

We welcome instructors and researchers from any discipline. Whether or not you have an engineering background, Aralia’s AI SDK (Software Development Kit) can help you find insights in the system:
1. Search suitable datasets across multiple Data Planets (e.g., SDG, UN…)
2. Align time and geography to return comparable tables
3. Let the LLM generate readable insight summaries
After this tutorial, you can complete a full run in Colab (no coding required). For advanced users, you can learn the five tool nodes and the LangGraph flow, and adopt the workflow in your classes or research.

📌 Tip: The clearer you define your research/decision question, the more precisely Aralia’s AI SDK will discover key insights.

After this tutorial, you will be able to:

Get insights by talking to AI
- Use the Aralia AI SDK to complete a semantic Q&A over data (with source lists and limitations).
Online Lab
- Run the full workflow in our carefully curated Colab environment (tweak parameters, customize prompts to see different behaviors and results).
Apply as a Lesson Plan
- Place problems into a “sample question bank” and re-run with one click (helpful in course use and repeatability).
Begin your Aralia AISDK journey
- Start building your own services with Aralia AI SDK.

What you’ll build

Minimum viable example (Malaysia GDP × Gini)
- Context:
  - In data analysis, we want to understand whether Malaysia’s GDP growth correlates with inequality (the Gini coefficient in SDG).
  - Gini coefficient:
    - A measure of inequality in the distribution of income or wealth, proposed by Italian statistician Corrado Gini in 1912.
    - Ranges from 0 to 1:
      - 0: perfect equality (e.g., everyone has the same income)
      - 1: perfect inequality (e.g., one person has all the income; everyone else has none)
- Question:
  - Is there a relationship between the average GDP growth at purchasers' prices from 2021 to 2024 and the Gini coefficient of each state in Malaysia in 2024?
- What the Agent does:
  - Searches Malaysia’s GDP and Gini in the SDG/UN Planets, extracts 2021–2024 and 2024 data → aligns by state → returns tables and an insight summary.
Pick your own (choose one from the “sample question bank”)
- Copy a complete prompt from the bank into Colab; or swap country/year to fit your topic.

Tools & resources you’ll use (no setup needed)

Colab Notebook:
- Includes environment setup, Secrets (LLM API Key, OAuth), one-click run, and output review.
Sample question bank:
- Ready-to-copy prompts.
Aralia OpenRAG SDK (advanced):
- Learn the five tool nodes and LangGraph flow;
- For advanced study, adjust the interpretation prompt or define custom nodes in the SDK.

How it works (understand the flow; no need to memorize)

Five tool nodes (chained with LangGraph):
1. Dataset search (semantic retrieval)
2. Analytics planning (decide x/y/filter and field formats)
3. Filter decision (fill in years/geography into query parameters)
4. Analytics execution (call Planet APIs, align time & geography)
5. Interpretation (LLM generates ≤300-word summary with caveats)
LLM choice:
- The Notebook uses your API Key to pick a provider (Google / OpenAI / Anthropic).
RAG & vectorization:
- Already handled by the system. Your job is to state the problem clearly.

graph LR
      A[Question] --> B["Aralia<br/>Search<br/>Node"]
      B --> C["Analytics<br/>Planning<br/>Node"]
      C --> D["Filter<br/>Decision<br/>Node"]
      D --> E["Analytics<br/>Execution<br/>Node"]
      E --> F["<br/>Interpretation<br/>Node"]
      F --> G[Answer]

subgraph "Workflow"
B
C
D
E
F
end

style A fill:#e1f5fe
style G fill:#e8f5e8
style B fill:#fff3e0
style C fill:#fff3e0
style D fill:#fff3e0
style E fill:#fff3e0
style F fill:#fff3e0

Ready? Let’s start with Step 1.

← Previous： Aralia AI SDK

Next： Quick Start →