Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/ksingh66/rideinsight

This is both a website and a standalone chatbot that is capable of taking CSV's and analyzing them to answer user questions.
https://github.com/ksingh66/rideinsight

ai chatbot csv django-application groq groq-api langchain llama3 pinecone

Last synced: 22 days ago
JSON representation

This is both a website and a standalone chatbot that is capable of taking CSV's and analyzing them to answer user questions.

Awesome Lists containing this project

README 

        
# RAG Chatbot for Ride Records Analytics



This project builds a **Retrieval-Augmented Generation (RAG) chatbot** that allows users to query ride records from a CSV (or text) file. The chatbot utilizes advanced natural language processing (NLP) and retrieval techniques to provide analytical insights based on the ride data, such as identifying top-performing chauffeurs, analyzing earnings over time, and answering other business-related questions.



The chatbot integrates **Pinecone** for vector storage and retrieval, **LangChain** for RAG workflows, and **Groq** for AI-based query processing.



The goal is to enable users to ask complex questions related to the ride data and receive insightful answers, including visualizations like graphs, trend analysis, and more.

## To use

- add 'rag_text.txt' in the same directory as main.py to add files to your vector storage.

- add 'bookings.csv' in the same directory as main.py to analyze csv files and ask the chatbot questions about your csv.

## Features



### Current Features

- **CSV/Text Parsing and Querying**: The chatbot can load and process ride data from `.txt` files, containing details such as `Booking#`, `Chauffeur`, `Price`, `Date`, etc.

- **RAG-based Query Responses**: The chatbot uses **LangChain** and **Pinecone** for context retrieval and to answer user queries dynamically.

  - Examples include:

    - "Who made the most money?"

    - "How many rides did Chauffeur X complete last month?"

    - "What is the total earnings for Chauffeur X?"

- **Data Retrieval**: The chatbot uses **Pinecone**'s vector search to retrieve relevant context from a seperate vector storage and then uses pandas to generate a summary of a CSV , then it generates answers using a **Groq** model through **LangChain**.

- **Flexible Data Input**: Ability to load ride records in text format and index them for later querying.



### Planned Improvements

## Top priority improvements

- Ride chat and Insight chat; Insight chat looks at multiple CSVs to perfrom a big picture analysis.

- Add most earned and least earned rides for each chauffeur.

- Add busiest to least busy days.

- Account for other CSV's that are not Transportation related; a general CSV analyzing AI.

- Make UI more interactive.

## Second priority improvements

- **Dynamic Graphing and Visualizations**: Integrating libraries like **Plotly** or **Matplotlib** to visualize ride trends, chauffeur earnings, and other key metrics.



### Technologies Used

- **Python**: Primary language for backend logic.

- **LangChain**: Framework for building RAG (Retrieval-Augmented Generation) workflows.

- **Pinecone**: Vector database for efficient document search and retrieval.

- **Groq (via LangChain)**: AI model for processing queries and generating responses.

- **Sentence-Transformers**: Used to generate embeddings for the ride records to store in Pinecone.

- **HuggingFaceEmbeddings**: For creating embeddings with HuggingFace models.

- **Plotly / Matplotlib** (future): For generating dynamic trend graphs and charts.