Custom AI Chatbot Avatar WordPress Plugin

Project Overview

The Custom AI Chatbot Avatar WordPress Plugin Creation Service is a web application that allows Vloe's clients to generate a customized WordPress AI chatbot avatar plugin for their websites. Each client can personalize the chatbot plugin by inputting their own Knowledge Base, Predefined Questions, Speaking and Idle Avatar Selection, and Language Preference.

The web application then generates a chatbot.zip file tailored to the client's needs, which can be deployed directly into their WordPress site. This improves customer engagement and reduces support inquiries by providing personalized, predefined questions, a knowledge base, language preferences, and avatars to address common customer inquiries.

Key Features

• Knowledge Base Customization: Clients can input specific knowledge related to their business or company for chatbot responses.
• Predefined Questions: Allows input of frequently asked questions for quick and automated responses.
• Speaking and Idle Avatar Selection: Clients can choose from a set of predefined avatars or upload their own images for the chatbot's speaking and idle states.
• Language Preference: Supports English and Canadian French. The chatbot's prompts and text-to-speech features will adapt based on the specified language in the shortcode.
• Custom Plugin Generation: Automatically creates a customized chatbot.zip plugin for WordPress.
• Shortcode-Based Deployment: Clients can manually specify the language using a shortcode and add the chatbot to specific pages.

Technical Implementation

The service is built with a PHP backend that handles data processing, avatar uploads, and dynamic plugin generation. The frontend uses HTML, CSS, and JavaScript for a simple and user-friendly interface.

The generated WordPress plugin integrates the OpenAI ChatGPT API for conversational responses and Google Cloud Text-to-Speech for voice interaction. The service is deployed as a dockerized serverless container function on Google Cloud Run, providing scalable, high-availability access.

The system architecture includes a web frontend for user input, a PHP backend for plugin generation, integration with Google Cloud and OpenAI APIs for TTS and translation, and a Dockerized deployment architecture.

My Contributions

As the team leader and scrum master, I led the development of the core functionalities, including:

• The front end in basic HTML and vanilla Javascript, and the back end using PHP development for user-uploaded keys and avatars and the entire code implementation.
• Dockerization of the web service and deployment as a serverless container to Google Cloud Run so that the service is accessible to everyone over the web as a Cloud Run service.
• Canadian French and English instruction page creation to demonstrate the instructions on how to create the chatbot using the service and an example workflow with template questions and knowledge base text.
• Transcript email generation when the user gives consent via an implemented checkbox in the frontend and provides their email before speaking to the chatbot.
• Avatar UX/UI improvement and speaking/idle state logic for determining when to display the idle animated chatbot with blinking or speaking animated chatbot with a moving mouth.

Team Members