🏠 Agentic AI Demo

Agentic AI walkthrough for HDB resale analysis → SQL → Data → Charts → Insights.

⚡ Tutorial Flow (4 steps)

1. Calling an LLM API

   • Setting up OpenRotuer API, ensuring dependencies are installed
   • 📄 File: intro_to_tool_calling.py

2. Generate SQL query output by passing in DB Schema

   • Instead of plaintext output, getting the LLM to give an SQL query
   • 📄 File: intro_to_tool_calling.py

3. Tool Time! Creating a tool to execute the above generated SQL query & give insights

   • Tool: execute_sql_query
   • 📄 File: intro_to_tool_calling.py

4. Adding a Visualisation tool + Putting it all together!

   • Add execute_visualisation_code tool & create plotly visualisations
   • Combine DB results + chart + analysis
   • 📄 Files:
     • final.py (Agent loop)
     • app.py (Streamlit UI)

5. Dockerising & Deploying to Airbase

   • Setting up Dockerfile
   • Linking repo to Airbase
   • Deploying to Airbase

🛠️ Setup

1. Clone the repo

git clone https://github.com/<your-org>/agentic-ai-demo.git
cd agentic-ai-demo

2. Create a virtual environment

python3 -m venv .venv
source .venv/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Add your OpenRouter API key

Create a .env file:

OPENROUTER_API_KEY=sk-your-key-here

---

🚀 Running the Final App

Locally with Streamlit

streamlit run app.py

👉 Go to http://localhost:8501

With Docker

docker build -t agentic-ai-demo .
docker run -p 8501:8501 agentic-ai-demo

---