Personal / Academic Build · 2025
API Recommendation System
Turning API selection into a guided workflow instead of a tab-heavy research exercise.
Role
Designer / Developer
Year
2025
Stack
- Next.js
- TypeScript
- OpenAI
- Vercel AI SDK
- MongoDB
- Tailwind CSS
Overview
The project looks at how a guided interface and LLM output can cut the time it takes to go from vague requirements to a shortlist of relevant API providers.
Challenge
- Make recommendations feel structured and explainable, not opaque.
- Handle provider data and user answers in a way that stays easy to scan.
- Design an AI workflow that earns trust through clarity.
Approach
- Combined the frontend tooling with structured AI output and analytics.
- Built a guided input flow instead of a single prompt box.
- Made the resulting recommendations comparable and useful.
Outcomes
- Built a working AI product with clear value for the user.
- Showed product thinking across UX, system design, and interface quality.
- Demonstrated how AI can support a decision without overwhelming the person making it.
Curious about the longer version, or the parts that didn't make the page? I'd be happy to talk it through. Get in touch