MACKENZIE CHATBOT

An accessible and intuitive helper for common I.T. problems.

​Chatbots saw a 92% use increase since 2019, making it the brand communication channel with the
largest growth. The #1 department supported by businesses who use chatbot technology is IT, at 53% of organizations. Chatbots have lot of benefits such as 24/7 availability and instant responses, and can benefit both users as well as businesses.

The MSU AI Club collaborated with the Wayne State Cyber Defense Club to create a Chatbot that aims to approach people to technology by providing answers to the difficulties they encounter in our virtualized world.

The Chatbot, Mackenzie, uses Natural Language Processing and Neural Networks to process queries from users and provide the most suitable intent from the JSON data. Mackenzie was programmed using the Python programming language as it contains many AI libraries. TensorFlow was used to build the Neural Networks, one model was built to predict category (password reset, online conference e.t.c) and one model was built per each category to determine tag number. Dataset was first created in google sheets and then transformed into JSON (JavaScript Object Notation) format. NTLK was used to pre-process text providing an array of text processing modules that are used for classification, tokenization, and stemming.

Mackenzie is deployed on a neutral website and is in beta stage.

🤖 Check it out: https://project-mackenzie-bot.herokuapp.com/