WhatsApp Chatbot for English Learners

Stack: RiveScript, MongoDB, WhatsApp Business API, Google Natural Language API, Google Translation API, Bad Word Filter API, Python

A WhatsApp chatbot for English learners to practice conversational English. The bot can detect languages, entities, bad words and sentiments to respond with proper words with the right tone. It can also loop in our staff if the bot isn’t able to respond.

Language Detection

Implemented Google Translation API to detect languages and encouraged the users to speak English with the bot.

Entity Detection

Utilized Google Natural Language API to detect entities such as names, addresses, activities for the bot to understand conversations and ask proper questions.

Bad Word Detection

Used Bad Word Filter API to encourage proper conversations.

User Sentiment Detection

Implemented Google Natural Language API to detect users’ sentiments and enable the bot to respond with proper tones.

Real Person Loop In

The chatbot is designed to respond to users in all possible scenarios, but in the case of bugs that the bot is not able to reply, the bot will send the conversation to our staff to loop in a real person in the conversation.

party conversation example

loop in a real person