AI & Vibe-Coding to unlock faster, freer, and more intuitive design.
Client
Vibe Coding
Year
2025
This self-initiated project explores how emerging AI tools and creative coding practices can elevate the design engineering process. Building on my experience with the Bellville platform, I revisited a set of features that were originally scoped but never implemented. Using Replit as my prototyping environment, I recreated this feature set from the ground up, integrating AI copilots, vibe coding principles, and rapid iteration techniques.
The goal was not only to bring a dormant idea to life, but to demonstrate a practical framework for embedding AI-assisted workflows into my broader design practice. From system architecture to front-end logic, every decision was informed by how AI can act as a collaborator, accelerating code writing, testing assumptions, and even inspiring unexpected creative solutions. This project serves as a prototype for my future approach to design engineering: deeply experimental, collaborative with machines, and rooted in user-centered intent.
Scope of Work
During the Bellville engagement, a powerful feature idea surfaced to enhance student dashboards by displaying real-world career data and support for career development. While it had stakeholder interest, the feature was never scoped due to resource limitations and shifting priorities. I saw this as the perfect sandbox to explore a few key questions:
Can AI tools enhance the design prototyping process—speed, creativity, and accuracy?
How close can a solo designer get to a complex working prototype using vibe coding and basic code knowledge?
Can AI-generated prototypes help win buy-in internally by making ideas tangible early?
I leaned into a designer-engineer mindset and used AI as both assistant and creative partner. My process looked like this:
Inputs from previous stakeholder interviews guided feature logic and experience flow.
Replit served as my live coding environment.
AI copilots (like ChatGPT and Replit Ghostwriter) helped me translate ideas into working code.
Vibe coding—a more intuitive, fast-feedback approach—let me iterate quickly without getting bogged down in syntax or boilerplate.
I integrated an external API to surface real-time career data for students, adding real-world utility.
The result was a fully functioning prototype deployed on Replit, featuring:
Real-time career data embedded into a student dashboard
Dynamic UI and logic modeled after our original Bellville platform architecture
A complete, testable experience that mirrored what we originally imagined—but never built
More importantly, the success of this solo effort generated real impact:
Internal buy-in: My prototype helped secure interest from leadership in integrating AI into our product development process.
Process innovation: The project sparked conversations about using AI-assisted prototyping for early-phase client engagements—saving time, reducing risk, and generating excitement.
Proof of concept: It validated that designers, even with minimal engineering experience, can use AI to build meaningful, functional tools—fast.
This project proved that AI and vibe coding aren’t just novelties—they’re powerful tools that can extend a designer’s reach into functional prototyping. With the right mindset, even complex ideas can be built and tested rapidly. For our team, this became a new model for what early product validation might look like.
Beyond the Bellville prototype, I also used this opportunity to experiment with building a small mobile game within Replit—just to test the limits of what I could create using AI prompting and limited coding experience. While unrelated to the core feature, it reinforced how flexible and powerful these tools are for designers who want to explore new mediums and skillsets.
This project became a personal milestone: a proof-of-concept for AI-augmented design, and a springboard for pushing creative boundaries in both product and play.
