Moll-E, An AI Personal Assistant
Overview
Inspired by movies like Iron Man and the desire to have an intelligent home assistant, we were passionate about taking this fictional technology into reality.Utilizing several AI models, we are able to dynamically record speech inputs, convert it to text, analyze the intent, and run commands based on the intention. For example, simple greetings are handled through the a simple chat bot while complex questions utilize AI search. Personal information can be stored on a databases and remembered, and events such as commands or alerts can be run using voice command. This Participated in and won 1st place in the Crimson Code hackathon, hosted by WSU, SEL, and Major League Hacking.
Features:
- Uses OpenAI’s GPT-3 API to generate natural language responses
- Utilizes Rasa NLU to understand and interpret user inputs
- Provides speech-to-text and text-to-speech functionality using OpenAI’s Whisper API
- Can be easily customized and extended with additional functionality
Built With
This section below contains the major frameworks/libraries used to bootstrap our project.
- Python
- spaCy
- Rasa NLU
- OpenAI
- gTTS
Getting Started
This project is functional but requires several API keys and a working microphone and speaker
Installation
- Get an API Key from OpenAI https://platform.openai.com
- Clone the repo
git clone https://github.com/Macbee280/CrimsonCode2023.git - Install the requirements.txt.
pip install -r requirements.txt
Usage
This program takes a while to initialize, follow the terminal outputs for further instructions
License
Distributed under the MIT License. See LICENSE.txt for more information.
Acknowledgments
Contributors on this project:
- Gabe DiMartino
- Nikita Dubinin
- Jonathan Simmons
- Dayna Roberts
Checkout Further Publications
