---
license: mit
title: Chiraag-s-Fine-Tuned-Classifiers
colorFrom: red
colorTo: red
sdk: docker
app_port: 8501
tags:
- streamlit
pinned: false
short_description: Gen-AI app to classify E-commerce & Banking customer queries
emoji: 🚀
sdk_version: 1.48.1
---

# 🏦🛒 Banking & E-commerce Customer Query Classifier  

## 📌 Overview  
This project demonstrates an **AI-powered text classification system** designed for the **banking and e-commerce domains**.  
Using a **fine-tuned DistilBERT model**, the system automatically classifies customer queries such as:  
- **Banking complaints** (e.g., credit card issues, loans, transactions)  
- **E-commerce queries** (e.g., delivery, returns, refunds, product complaints)  

The goal is to **streamline support processes** by routing queries to the correct department, improving resolution speed and customer satisfaction.  

---

## ⚡ Key Features  
- 🔹 Fine-tuned **DistilBERT** for domain-specific classification  
- 🔹 Supports **multi-category classification** across banking & e-commerce  
- 🔹 **Offline deployment** for fast and secure query handling  
- 🔹 Interactive **Streamlit interface** for real-time testing  
- 🔹 Can be upgraded for **enterprise-level applications**  

---

## 🛠️ Tech Stack  
- **Model:** DistilBERT (fine-tuned)  
- **Frameworks:** Hugging Face Transformers, PyTorch  
- **Interface:** Streamlit  
- **Deployment:** Hugging Face Spaces using Docker containers  
- **Cloud & Compute:** Google TPU 

---

## 🚀 How It Works  
1. User enters a customer query or complaint.  
2. The model classifies it into a **specific category** (banking or e-commerce).  
3. The system returns the predicted category for **automated query routing**.  

---

## 📊 Example Categories  
### 🏦 Banking  
- Credit Card Issues  
- Loan & EMI Queries  
- Transaction Failures  
- Account Services  

### 🛒 E-commerce  
- Delivery Issues  
- Returns & Refunds  
- Payment Problems  
- Product Complaints  

---

## 💡 Use Cases  
- 🏦 **Banking:** Faster resolution of customer complaints  
- 🛒 **E-commerce:** Automated support ticket classification  
- 🤖 **AI Agents:** Integration into chatbots or virtual assistants  

---

## 📂 Project Structure  

Chiraag-s-Fine-Tuned-Classifiers<br>
│<br>
├── bank_customer_ticket_category_classifier/      # Banking complaint classifier model or code<br>
├── E-commerce-customer-query-classifier/         # E-commerce query classifier model or code<br>
├── .gitattributes                                # Git configuration file<br>
├── bank-accuracy.png                             # Banking model accuracy plot/image<br>
├── e-comm-accuracy.png                           # E-commerce model accuracy plot/image<br>
├── Dockerfile                                    # Container build settings<br>
├── README.md                                     # Project documentation<br>
├── requirements.txt                              # Python dependencies<br>
└── streamlit_app.py                              # Streamlit web application<br>


---

## 🧑‍💻 Author  
**Chiraag P V**  
AI Professional specializing in **LLMs, Generative AI, and Text Classification**.  
Experienced in building **scalable AI solutions** with LangChain, LangGraph, Azure, and NVIDIA GPUs.  

---

## 🌟 Try it Out  
🚀 Live Demo on **Hugging Face Spaces**:  
👉 [Chiraag-s-Fine-Tuned-Classifiers](https://huggingface.co/spaces/Chiraag-P-V/Chiraag-s-Fine-Tuned-Classifiers)