AI platform / Web development
Bundance AI
Turning a shopping idea into something you can explore.
Bundance brings product discovery and an AI shopping assistant into one experience. I built the web platform around a simple journey: describe what you need, explore relevant products, and keep the options worth coming back to.
- Focus
- AI product discovery
- Platform
- Web application
- Built with
- React · Next.js · TypeScript
The challenge
Shopping rarely starts with a perfect product name. Someone might be choosing a retirement gift, comparing a few possibilities, or trying to stay within a budget. A useful interface has to carry that context into the results instead of making every refinement feel like a new search.
For Bundance, the design and development challenge was to connect a conversational experience to concrete product information. The assistant needed room to explain and refine, while product images, prices and merchant links still needed to be easy to scan. Neither part could become an obstacle to the other.
The approach
A conversation beside the catalogue
The interface puts product discovery and the assistant together. Shoppers can review the results while continuing the conversation, which gives follow-up requests a clear visual context. The layout treats the product list as the working surface and the assistant as a way to narrow the choice.
Product cards with a clear next step
Images, titles, prices and a visible product action give each result a consistent structure. Wishlist controls separate an item worth saving from an item ready to visit. That distinction supports browsing without forcing an immediate purchase decision.
A connected application, beyond the landing page
The React and Next.js application includes shopping conversations, account areas and a wishlist. TypeScript supports shared application models, while server-side routes handle the integrations behind the interface. Keeping these responsibilities separate makes the product experience easier to extend as discovery needs evolve.
03 / Inside the project
The technology behind Bundance AI
The languages and tools used to bring the project to life.
The stack
- React & Next.js
- Shopping interface and application routes
- TypeScript
- Shared application models and product logic
- PostgreSQL & Prisma
- Application data and persistence
- Python
- Supporting data and AI processing tools
Code by language
- TypeScript
- 78.1%
- Python
- 9.5%
- Vue
- 3.8%
- CSS
- 3.6%
- Other
- 4.9%
The result
The result is a shopping workspace that connects natural-language discovery with browsable products and a personal shortlist. It demonstrates the full application work behind an AI interface: navigation, product presentation, account features and the paths between them.
For a similar AI web application, the useful starting point is the decision a person needs help making. The interface, data model and assistant behaviour can then support that decision together. Bundance is an example of that approach applied to shopping and gift discovery.
Have something in mind?
Let’s build your next project.
Tell me what you want people to do, and we can work out the right experience to support it.
