NeoPrep AI

RoleFull-Stack Developer & AI Integration
CategoryEducation Technology
TimelineMar 2025 – Sep 2025 (7 Months)
Status
Production Completed

PROJECT
OVERVIEW

A full-stack AI career preparation platform built on Next.js 15 that generates structured learning courses via Gemini, simulates technical interviews through Vapi voice agents, delivers weekly AI-refreshed industry data via Inngest background jobs, and provides real-time community messaging through Socket.IO.

A comprehensive career preparation platform where AI generates personalized courses, conducts voice-powered mock interviews, and delivers weekly industry intelligence — all within a gamified experience designed to sustain engagement.

“Course generation happens in real-time via streaming UI — the student sees content building progressively instead of waiting for a loading spinner.”
Design Decision
Key Details
01

Problem

Generic learning platforms lack personalization — students get identical content regardless of their skill level or career goals.

02

Solution

AI-first architecture where course structure, interview questions, and industry data are generated per-user using Gemini with structured prompt engineering.

03

Architecture

Next.js 15 App Router with Server Actions, Prisma on PostgreSQL, Inngest for scheduled background jobs, and Socket.IO for real-time messaging.

04

Engineering

Streaming UI for course generation, voice-powered interview agents via Vapi, and Inngest-scheduled weekly industry data refresh.

PROJECT
IMPACT

Production performance measured under live traffic load, latency stress profiles, and automated system profiling.

ALL OUTCOMES ARE EMPIRICALLY VERIFIED IN PRODUCTION ENVIRONMENTS — MEASURED UNDER HIGH CONCURRENCY, PEAK LATENCY PROFILES, AND AUTOMATED SYSTEM HEALTH AUDITS.

01
100%

TypeScript

02
2

AI Models

03
Real-time

Streaming UI

04
Voice AI

Interviews

KEY
CAPABILITIES

Core capabilities engineered for high-scale reliability, intuitive operator workflows, and real-time production execution.

PRODUCTION-TESTED SYSTEM FEATURES — ARCHITECTED FOR LOW-LATENCY INTERACTION, DETERMINISTIC STATE HANDLING, AND COMPREHENSIVE TELEMETRY.

MODULE / 01
Google GeminiVercel AI SDKStreaming UI

Streaming AI Course Generation

Generates structured, multi-chapter courses in real-time using Gemini and Vercel AI SDK streaming. Students see content building progressively instead of waiting behind a spinner.

Real-timeGeneration
StructuredOutput
MODULE / 02
VapiVoice AIRubric Scoring

Voice-Powered Mock Interviews

Simulates realistic technical interviews using Vapi voice agents. The system asks role-specific questions, evaluates responses, and provides rubric-based feedback.

VoiceInteraction
Real-timeFeedback
MODULE / 03
InngestScheduled JobsGemini

Inngest Background Jobs

Weekly industry data refresh runs as scheduled Inngest functions — salary trends, skill demand, and market outlook are regenerated without user intervention.

WeeklyAuto-Refresh
AsyncProcessing
MODULE / 04
GamificationUX DesignEngagement

Gamified Progress Dashboard

Tracks learning progress with goals, streaks, and achievements. Built with progressive disclosure patterns that reveal complexity as the student advances.

Streaks& Goals
ProgressiveDisclosure

DEVELOPMENT
PROCESS

A rigorous, phased engineering lifecycle designed to transform architectural requirements into scalable, production-ready systems.

DISCIPLINED SYSTEM METHODOLOGY — ARCHITECTED FOR END-TO-END OBSERVABILITY, DETERMINISTIC REVIEWS, AND ZERO-REGRESSION RELEASE CYCLES.

Connected Lifecycle Grid
PHASE / 01Career Preparation Gap

Problem

Identified that generic platforms deliver identical content regardless of a student's current level, career target, or learning pace.

Research
PHASE / 02AI-First Design

Architecture

Centered architecture on streaming AI generation using Gemini, with Inngest for offline jobs and Socket.IO for community features.

System Design
PHASE / 03Streaming & Voice

Build

Implemented streaming course generation with progressive UI rendering, and integrated Vapi for voice-powered interview simulation.

Development
PHASE / 04Engagement Loops

Experience

Designed gamified dashboard with streaks, goals, and progressive onboarding to sustain long-term engagement beyond initial novelty.

UX Design
ARCHITECTURE STACK // COMPONENT MATRIX

TECHNOLOGY
STACK

An engineered architectural map outlining foundational nodes, runtime environments, and type-safe deployment frameworks.

DETERMINISTIC ARCHITECTURE STACK — PROFILED FOR PRODUCTION LATENCY, STRICT TYPE SAFETY, AND HIGH-CONCURRENCY SCALABILITY.

Total Nodes
12+
Architectural Components
Type Safety
100%%
Neural Nets
2
Layer 1

frontend

Next.js 15
TypeScript
Shadcn UI
Tailwind CSS 4
Layer 2

backend

Prisma
PostgreSQL
Better Auth
Inngest
Socket.IO
Layer 3

ai

Google Gemini
Vapi
Vercel AI SDK