# CDA Bus Route Assistant using Process Mining and LLMs
## Overview
This project combines process mining techniques with large language models to build a conversational assistant for the CDA Islamabad Bus Service.
Instead of manually searching bus schedules, users can ask natural language questions about routes, travel time, transfers, and departures.
The assistant retrieves relevant route information from structured transit data and generates responses using the Groq API with Llama 3.
---
## Features
- Natural language bus route search
- Direct route discovery
- Transfer route planning
- Travel time estimation
- Last bus lookup
- Route search by stop
- Fuzzy stop name matching using RapidFuzz
- LLM-generated responses using Groq
---
## Technologies
- Python
- Pandas
- RapidFuzz
- Groq API
- Llama 3
- Process Mining Event Logs (XES)
---
## Project Files
task1.py Process mining task
task2.py Process mining analysis
task3.py Additional process mining task
task5.py Intelligent bus route assistant
routes.csv CDA bus route dataset
cda\_log.xes Event log
---
## Installation
pip install -r requirements.txt
Create a .env file:
GROQ\_API\_KEY=YOUR\_API\_KEY
---
## Run
python task5.py
Example queries:
How do I get from NUST Metro Station to Blue Area?
What is the last bus from Faiz Ahmed Faiz Metro Station?
Which routes pass through G-10?
---
## Course
Process Mining
