Agentic AI in Action: Exploring AI Agents & Tool Calling

Organized by: ACM-W

This repository contains the complete code used in the ACM-W workshop “Agentic AI in Action: Exploring AI Agents & Tool Calling”.

In this workshop, students learn the basics of:

• AI Agents
• Tool Calling
• Retrieval-Augmented Generation (RAG)
• MCP (Model Context Protocol)
• LangChain & LangGraph

Students also build a working AI assistant that can:

• answer questions from uploaded documents
• access external web information
• perform calculations using tools

---

Basics of Agentic AI

Traditional AI chatbots only generate answers from what they already know.

An Agentic AI system can:

• decide which tool to use
• retrieve information dynamically
• access external sources
• perform actions/calculations
• generate more accurate responses

This makes the AI assistant:

• More intelligent
• More flexible
• More interactive
• More accurate for real-world tasks

---

What This AI Assistant Does

• Allows users to upload TXT or PDF documents
• Uses RAG to answer questions from uploaded files
• Uses a Web Search Tool for external/live information
• Uses a Calculator Tool for calculations
• Uses MCP for tool communication
• Uses LangGraph for workflow orchestration
• Routes queries intelligently to the correct tool
• Generates grounded responses
• Maintains conversational interaction during the session

---

Tech Stack

• Python
• Streamlit
• LangChain
• LangGraph
• MCP
• ChromaDB
• NVIDIA NIM
• Requests
• PyPDF2
• python-dotenv

---

How to Run the Project

Clone the github repository

git clone <PASTE YOUR GITHUB REPO LINK HERE>

Windows

python -m venv venv
venv\Scripts\activate

Mac/Linux

python3 -m venv venv
source venv/bin/activate

Install all required python packages

pip install -r requirements.txt

Create a file to store your api keys (.env) & add your api keys

NVIDIA_API_KEY=nvapi-xxxxxxxxxxxxxxxx

TAVILY_API_KEY=tvly-xxxxxxxxxxxxxxxx

Run the streamlit application

streamlit run main.py

Important Notes:

• The vector database is created locally on your system
• Users can upload multiple files
• The AI assistant dynamically selects tools depending on the query
• Web queries require internet access
• Do NOT upload your .env file or API keys to GitHub
• To reset everything, delete the chroma folder or use the clear option if available