README.md

---
license: mit
language:
- en
pretty_name: Wireless plans
size_categories:
- n<1K
---
---
license: mit
language: 
- en
tags:
- telecom
- mobile-plans
- RAG
- question-answering
---

# 📊 Telecom Plans Dataset (Bell, Virgin Plus, Lucky)

This dataset contains ~130 curated mobile plan entries from **Bell**, **Virgin Plus**, and **Lucky Mobile**. It was extracted from provider PDFs and normalized into a structured CSV for use in **RAG-based Question Answering** systems.

## 📂 Dataset Structure

Each row corresponds to a unique plan offering and contains:

- **Provider**: Carrier name (Bell, Virgin Plus, Lucky)  
- **Plan Name**: Marketing name (e.g., *Elite Lite (New)*, *BYOP 60GB*)  
- **Price**: Monthly price in CAD  
- **Data**: Data allowance (e.g., 20GB, 50GB, Unlimited)  
- **BYOD**: Boolean — `True` if BYOD plan, else device financing  
- **Contract**: Contract type (e.g., SmartPay, 24-month)  
- **Hotspot**: Boolean — whether hotspot is included  
- **Notes**: Additional info (New, Add-a-line, HUG, Student, Region, etc.)  
- **Description**: created using python function - language summary for use in RAG  

### Example Row
```csv
Provider,Plan Name,Price,Data,BYOD,Contract,Hotspot,Notes,Description
Bell,Elite Lite (New),60,60GB,False,SmartPay,True,New,"Bell offers the 'Elite Lite (New)' plan for $60/month. It includes 60GB of data, device financing over 24 months, hotspot support, and is available for new customers."
```

### Usage
```python
from datasets import load_dataset

dataset = load_dataset("Sathya77/telecom-plans")
print(dataset["train"][0])
```

💡 Applications

Retrieval-Augmented Generation (RAG) demos

FAQ-style question answering

Plan recommendation assistants

Semantic search experiments

📊 Notes

Plans are based on publicly available provider documents (as of 2025).

This dataset does not reflect live pricing.

Region-specific entries are based on plan names (e.g., Ontario, Quebec).