Overview
An experimental project built purely to understand conversational AI — how to structure prompts, maintain context across a conversation, and give a chatbot something resembling a consistent personality.
The Challenge
Off-the-shelf chatbot demos tend to feel robotic or forget context after a couple of exchanges. The challenge was making a lightweight front-end feel more natural without a heavy backend.
Approach
- Explored prompt engineering techniques to give the AI consistent tone and memory of recent messages.
- Built a simple, chat-style front-end focused entirely on the conversation, with no unnecessary UI noise.
- Iterated on response formatting to make replies feel more natural and less like raw API output.
The Solution
A playful conversational AI interface that demonstrates practical prompt-engineering and front-end integration with LLM APIs.
Outcome
This project deepened my understanding of AI integration patterns that I later reused in more serious tools, like the A&K Study Helper AI.