Architected and led development of a scalable exam-prep platform with real-time analytics, AI tutoring, and cross-platform apps.

Whiz Study is a full-scale exam-prep platform for SAT, PSAT, and AP courses serving 60,000+ students, with 6,000+ daily active learners.

When I joined as the technical lead, the goal was to transform an early prototype into a scalable, production-grade platform with:

Powerful topic-based practice

An interactive AI tutor

Full analytics pipeline

Mobile + Web apps

High-availability backend

Fast content tooling for educators

Over 18 months, I architected, built, and scaled the entire system from the ground up — spanning backend, frontend, mobile, infra, analytics, and team leadership.

The platform now supports tens of thousands of students, powers a complete practice experience, and is actively used by schools, teachers, and self-studying students.

Engineering Manager for 6 developers

Lead designer of system architecture & database

AI Tutor development + LLM integration

Led mobile app development (React Native/Expo)

Managed production deployments & scaling

Coordinated product roadmap with founders

Introduced code standards, CI/CD, and monitoring

Multi-app monorepo (Web, Mobile, Backend, Analytics)

Full database & schema design using PostgreSQL + Prisma

Real-time analytics service (TimescaleDB + RabbitMQ)

SAT/PSAT practice experience (quizzes, tagging, difficulty)

AI Tutor simulation + LLM integration (OpenAI)

AP digital practice exams system

Performance optimizations for 6k DAU

Custom CMS integration for content teams

Mobile app (iOS + Android) using Expo Router

Automated scoring, insights, and progress tracking

Scalable infrastructure on Vercel + Docker services

Role-based dashboards for students, teachers, admins

Frontend: Next.js, React, Typescript, Tailwind Mobile: React Native, Expo, Expo Router Backend: Node.js, Fastify, Express, Prisma Database: PostgreSQL (Supabase), TimescaleDB Messaging: RabbitMQ Infra: Vercel, Docker, Hetzner, PM2 AI: OpenAI, custom prompt engine Analytics: Timescale queries, real-time event collection Payments: Stripe (subscriptions), RevenueCat (mobile)

18 months Architecture → Platform build → Scaling → AI tutor → Full product launch

40% faster onboarding after UI redesign

AI tutor increases completion rate of practice sets

Mobile app adoption growing ~20% month over month

Zero major downtime during peak exam seasons

Successfully replaced multiple 3rd-party tools with in-house features

Improved analytics accuracy & reduced data-related bugs