Skip to content

Latest commit

 

History

History
277 lines (145 loc) · 13.5 KB

README.md

File metadata and controls

277 lines (145 loc) · 13.5 KB
Raw

Karna Bot

Smarter , Flexible, State-Aware Conversation Chat Bots for Your Quick Services

License: MIT build-status workflow Contributors workflow GitHub issues

Karna is a Chatbot designed to facilitate ride booking on the Namma Yatri Ride Hailing Platform. The chatbot is integrated with WhatsApp and provides multilingual support to users. Its primary goal is to deliver a seamless and convenient booking experience, free of any complications. Additionally, Karna offers a range of other features to enhance user experience.

Namma Yatri Hackathon Winner 🏆

This project was selected as the Grand Prize Winner in the Namma Yatri Open Mobility Hackathon , Among 170,000 participants.

Try It Out

Experience the convenience of seamless communication with Karna, our WhatsApp bot! Simply scan the QR code below or click on this Link and embark on a journey of effortless interaction.

Features

System Features

  • Multi-Platform Support: The Karna Chat bot is designed to be integrated with multiple platforms, including WhatsApp, Telegram, Slack, and others.

  • Parallel Conversation Support: The bot is highly scalable and can handle a large number of conversations for multiple users parallel.

  • Conversation Isolation : The bot is designed to handle multiple conversations at the same time without interfering with other user conversations.

  • State-Aware Conversation: The bot is designed to maintain the state of the conversation with the user. This means that the bot can remember information provided by the user earlier in the conversation and use it to provide more personalized responses. For example, if the user has already provided their location, the bot can use that information to suggest nearby ride options.

  • Integrated Template Engine: The bot includes a template engine that allows the generation of messages and responses to user queries quickly and efficiently in multiple languages.

  • Horizontal Scalability: The bot is designed to be horizontally scalable, meaning that it can handle a large number of users by distributing the workload across multiple servers.

  • Auto Conversation Closing: The bot is designed to automatically close conversations with users after a specified period of inactivity.

Bot Features

  • End-to-End Ride Booking : Users can book a ride in Namma Yatri Platform from start to finish using the Karna Chat bot.

  • Ride Status Tracking : The bot provides users with real-time updates on the status of their ride, such as driver ETA and more.

  • Ride Cancellation : Users can cancel their booked ride using the bot.

  • Feedback Provision : The bot allows users to provide feedback on their ride experience.

  • Multi-Language Support : The bot supports multiple languages to cater to a wider user base, currently Kannada, Hindi and English are supported.

  • Starred Places Management : Users can manage their frequently used or favorite locations and use them for booking rides quickly.

  • View Past Rides : Users can view their ride history and check ride details.

  • Namma Yatri Open Data : The bot provides access to Namma Yatri Open Data.

  • Additional Features : The bot provides additional features such as support, FAQ section, and more.

Tech Stack

  • Camunda: is a popular open-source platform for workflow automation and business process management.

  • Spring Boot: is a popular framework for building REST APIs in Java.

  • Redis: is an in-memory data structure store that is often used as a database, cache, and message broker.

  • Mockoon: is an open-source tool that helps developers to simulate APIs. Mockoon was used to simulate Namma Yatri APIs.

  • MongoDB: is a popular NoSQL document database. We used MongoDB, to store our system's data and analytics details.

By using open-source technologies, we were able to build a scalable and robust ride-booking system that is both efficient and cost-effective.

Bot In Action

Watch the below video to get a better idea on how bot works

Krana Bot Demo

System Architecture Overview

The Karna Chat bot is a system composed of multiple components, each responsible for a specific function in the overall architecture. The following are the key system components:

  • Message Platform : This component acts as the primary interface between the user and the bot. Users can send messages or interact with the bot through various platforms such as WhatsApp, Slack, or Telegram.

  • Message Gateway : This component receives messages from the message platform and forwards them to the bot engine. It is responsible for handling platform-specific logic and converting it to a generic format that can be understood by the bot engine. For the WhatsApp platform integration, we have used the gupshup.io platform.

  • Workflow Orchestrator : This component manages the workflow of the bot and persists conversation state. It is responsible for deciding the next state of the conversation based on the user's input. We have used the Camunda BPMN engine for this component, which has its own relational database to store the state of the conversation.

  • Template Engine : This component generates messages and responses to user queries quickly and efficiently in multiple languages.

  • Incoming Message Handler : This component handles incoming messages from the Message Gateway.

  • Backend Event Handler : This component handles backend events from the service provider platform, such as Namma Yatri, to receive users' ride updates.

  • Document DB : his component maintains user configuration and template configuration and helps in the analytical dashboard.

  • Caching Layer : This component caches user and template context to provide a faster response to the user.

  • Live Analytics Dashboard : This component provides insights into the usage of the Karna Chat bot, such as the number of users, user trends, and active conversations. This information can be used to improve the bot's performance and user experience.

  • Namma Yatri Backend Mock Server : his component is responsible for simulating the backend APIs of the service provider platform, Namma Yatri, and providing the simulated data to the bot. The mock server provides an interface that mimics the behavior of the actual backend APIs, allowing the bot to interact with it as if it were interacting with the real APIs

  • Backend Event Simulator Host : This component is responsible for simulating the backend events from the service provider platform, such as Namma Yatri, and sending the simulated events to the bot. The simulator host provides an interface that mimics the behavior of the actual backend event streams, allowing the bot to receive the simulated events as if they were real. In this case we simulated ride updates event of Namma Yatri.

  • CI / CD Pipeline : to deploy the code to the cloud and manage the deployment and versioning of the code.

Low Level Design

The low-level class diagram illustrates the implementation details of the Namma Yatri chatbot, including the classes and their relationships.

Technical Documentation

View Detailed Technical Documentaion

Prerequisites to use Bot

To use Karna bot, you need to have an active WhatsApp account and a smartphone with a reliable internet connection. Ensure that your WhatsApp account is registered with the phone number you wish to use to interact with the Namma Yatri Service.

Open Analytics Dashboard

The analytics dashboard provides insights into the usage of the Karna Chat bot, such as the number of users, user trends, and active conversations. This information can be used to improve the bot's performance and user experience.

View Live Analytics DashBoard

Roadmap & Future Scope

  • Live Analytics Dashboard.
  • Replacing mock server with Namma Yatri server SDK integration
  • Sharing driver’s live location as message to end users after ride assignment.
  • Maintaining conversation history for users.
  • Adapter design pattern for messaging client to easily integrate with multiple message platforms.
  • Building admin dashboards to manage templates and conversation flow.
  • Integration of alerting tool for app exceptions alerts.
  • Integrate with PostHog for product analytics, user analytics and more.
  • Manual takeover of bot replies and handling it by humans.
  • Campaign management using Karna bot.
  • Multi tenant support for using Karna bot.

How to set up

This guide will walk you through the process of setting up the Karna Chatbot on your local machine. Please make sure that you have the following prerequisites installed before proceeding:

  • Node.js (v14 or higher)
  • Java 17 or higher
  • Docker (v20.10 or higher)
  • Docker Compose (v1.29 or higher)

Step 1: Clone the repository

The first step is to clone the Karna ChatBot repository from GitHub:

git clone git@github.com:nishanb/Namma-Bot.git

Step 2: Setup Gupshup and get API Credentials

Follow this doc to set up WhatsApp access api on GupShup

Step 3: Setup Mock API Server

Follow this doc to set up Mockoon

Step 4: Setup Event Simulator Service

Follow this doc to set up Ride Event Simulator

Step 5: Setup MongoDB

Follow this doc to set up MongoDB

Step 6: Setup Karna Bot Env File

Create a .env file in the root directory of the project and add the following environment variables:

# Gupshup API Credentials
GUPSHUP_API_KEY=
GUPSHUP_APP_NAME=
GUPSHUP_HOST=https://api.gupshup.io/
GUPSHUP_SOURCE_CONTACT=
# Mongo Config
DB_URI=
# Mock Api Server config
BACKEND_HOST=http://localhost:3000/api/v1
# Event simulator host config
EVENT_SIMULATOR_HOST=http://localhost:3001
# Camunda config (used by cpo)
CAMUNDA_PASSWORD=secret
## Kana bot host config  
CALLBACK_HOST=http://localhost:8000

Step 7: Setup & Deploy Karna Bot

To deploy an KarnaBot using docker compose you can run the following:

sudo docker-compose --env-file ~/.env up --build
  • Now to access the camunda dashboard you can go to http://localhost:8080. The default username is demo and password is demo.
  • Send a message to registered WhatsApp number to start the conversation with Karna Bot.

License

This project is licensed under the MIT License. You can find the license file in the root directory of this repository.

Contribution

We welcome contributions to this project. If you find a bug or would like to suggest a new feature, please create a pull request or submit an issue. Before submitting your pull request, make sure to run the tests and ensure that they pass. We also ask that you follow our coding guidelines.

Issues

If you encounter any issues while using this project, please create a new issue on our GitHub repository. We will do our best to address the issue as soon as possible. When creating an issue, please provide as much detail as possible, including steps to reproduce the issue and any error messages you received.

Conclusion

In conclusion, the ChatBot solution presented here provides an efficient and user-friendly way for interacting with various services. By leveraging popular messaging platforms like WhatsApp, the chatbot enables a seamless conversation-based experience for users to perform different actions without leaving the chat interface. The chatbot solution also includes various components such as a message platform, message gateway, workflow orchestrator, template engine, incoming message handler, backend event handler, conversation process orchestrator, document database, and more. These components work together to provide a robust and scalable architecture for handling large volumes of user interactions.

We welcome any feedback or suggestions you may have to help improve the chatbot solution further. With the ability to integrate with different messaging platforms and services, this solution can be adapted to various use cases and industries. As the chatbot ecosystem continues to grow, we believe that solutions like this will play an increasingly important role in providing seamless and efficient interactions between users and services.