Case Study
PulsePrep
An AI-assisted nursing licensure review platform built around adaptive exams, learner analytics, and progress visibility.
The Problem
Nursing licensure review often depends on static question banks, manual progress tracking, and disconnected study workflows. Students need practice, but instructors and reviewers also need visibility into weak areas, progress, and engagement.
- Practice exams needed to feel closer to the actual PNLE review flow.
- Students needed feedback beyond a final score.
- Review operations needed analytics without manually consolidating spreadsheets.
- Question generation and study workflows needed to scale without constant manual content work.
The Solution
PulsePrep combined adaptive exams, AI-generated questions, learner progress tracking, and analytics into one review platform. The goal was not just to digitize quizzes, but to help students and reviewers understand performance over time.
Adaptive Exams
Built exam workflows that supported structured practice, progress visibility, and repeated review sessions.
AI Question Support
Integrated AI-assisted question generation to reduce manual content creation and support review coverage.
Learner Analytics
Tracked progress, exam history, and weak areas so students could see where to focus.
Reviewer Visibility
Created dashboards to make performance and engagement easier to monitor across learners.
Production Delivery
Owned architecture, database design, API workflows, deployment, and product iteration.
Fast Iteration
Used Nuxt.js and Supabase to move quickly from prototype to usable product while keeping the stack small.
Architecture
The platform used Nuxt.js for the application layer, Supabase for authentication and data workflows, PostgreSQL for structured exam and learner data, and OpenAI API integration for AI-assisted content workflows.
- Nuxt.js handled the learner and admin experience.
- Supabase supported authentication, storage, and database-backed product workflows.
- PostgreSQL modeled exams, attempts, questions, results, and progress tracking.
- OpenAI API supported AI-assisted question generation and content workflows.
Outcome
PulsePrep reached a peak of 250+ active users and validated demand for a more adaptive, analytics-driven review experience. The product is now inactive, but it remains an important example of turning a learning workflow into a production platform.
- Reached 250+ active users at peak.
- Combined exams, analytics, progress tracking, and AI workflows in one product.
- Reduced manual review operations by making student progress easier to see.
- Clarified the product and architecture tradeoffs behind AI-assisted education tools.
What I Learned
- AI features need strong product constraints. Generation alone is not the product.
- Education tools are only useful when they improve feedback loops for learners and reviewers.
- Analytics matter most when they help users decide what to do next.
- A small stack is an advantage when the product still needs fast iteration.
Interested in education or AI-enabled product work?
I build practical products around real workflows, clear constraints, and measurable outcomes.