Excel Worker ReAct (Reasoning and Acting) AI Agent using LangChain

This Agent enables analysis of Excel files using free-form queries as user input.

Note: This is an older LangChain-based implementation. I recommend re-creating it using LangGraph react agent or checking out this Google ADK implementation.

If you have any questions or would like to collaborate, feel free to reach out to me on LinkedIn. You're more than welcome!

Core Architecture

• Built with the ReAct (Reasoning and Acting) agent pattern using LangChain
• Uses GPT-4o as the underlying LLM
• Integrates DuckDB and Pandas for structured data processing

Main Tools

• preview_excel_structure: Analyzes the file structure and data types
• complex_duckdb_query: Handles complex SQL operations (grouping, aggregations)
• simple_dataframe_query: Executes row-level operations using Pandas

Key Features

• Robust error handling and state management
• Intelligent handling of NULL/empty values
• Automatic data preprocessing
• Supports complex SQL queries with WITH clauses
• Built-in data visualization capabilities
• Very well decorated print line allowing to follow the execution logic

ReAct pattern guides the workflow

• Input: Excel file and a free-form user query.
• Intermediate Step: The agent determines which tools to call and the best approach to answering the original user question.
• Output: Execution result from the LLM generated query, answering the user query.

Intallation

Prerequisites

• Access to JupyterLab, Google Colab, or another interactive computing environment to run this Jupyter Notebook.
• Access to LLM API.

Step 1: Clone the Repository

Clone this repository to your local machine:

git clone <REPOSITORY_URL>
cd <PROJECT_FOLDER>

Step 2: Open Jupyter Notebook in JupyterLab

Ensure that <PROJECT_FOLDER> is accessible in JupyterLab by setting it as your working directory in JupyterLab.

• In JupyterLab, use the "Open from Path" option to load ExcelWorkerReActAgent.ipynb.
• Similarly, load .env and populate the variable keys with appropriate values.
• The first cell in the Notebook installs the required libraries: %pip install langchain langgraph pandas python-dotenv duckdb

Step 3: Run the Jupyter Notebook

To execute the notebook, select each cell and press Shift + Enter.