The Problem
Our client approached us with an AI platform for use in therapy. Only, their previous development team had abandoned the project, leaving it unfinished and broken. Leveraging our knowledge of artificial intelligence and secure back-end development, we happily stepped up to the challenge.
Our Solution
In taking over our client’s project, we first thoroughly reviewed their existing progress and assessed the level of work required to successfully integrate complex AI functionality. Once we caught up, we built the client a secure back end based on ChatGPT that enables users to talk with their chatbot using natural language. Crucially, through prompt engineering, we’ve also allowed our client to alter the personality and purpose of their chatbot, simply by updating a prompt in the back end.
What We Did
Back-End ChatGPT Interface
We created a back-end interface based on the existing user experience provided by ChatGPT. This creates a familiar, yet effective, user-friendly interface that allows our client and its users to train the AI chatbot with ease. Unlike traditional ChatGPT integrations, however, we add an extra layer of security through Azure Cloud Services and our standardized development procedure.
Prompt Engine Configuration
By accessing the platform’s back end, our client can easily reconfigure the AI chatbot using a prompt. For example, with prompt engineering, the client can tell the system to “talk like a pirate” and then, when users chat with the bot, the ChatGPT-based chatbot will attempt to mimic the tone of a pirate.
Professional Overhaul
As a rescue project, our client’s platform needed a vast overhaul to ensure it was both secure and functional. We provided a comprehensive professional overhaul, tying up loose ends and refining the AI chatbot’s functionality.