SynapQ
An intelligent, text-based interview preparation platform powered by Active Recall & Spaced Repetition (SM-2).
SynapQ is a focused, text-based interview preparation system built on proven cognitive science principles: Active Recall, Spaced Repetition (SM-2), and simulated pressure. Designed to help candidates retain deep technical & behavioral knowledge and express it clearly under time constraints without relying on bloated voice AI.
The Problem
Traditional interview preparation is passive and ineffective: candidates read answer banks without active recall, forget information quickly due to the Ebbinghaus forgetting curve, and experience severe anxiety under real interview time constraints.
The Solution
Built a full-stack platform combining a customized SM-2 Spaced Repetition algorithm that schedules reviews precisely before forgetting occurs, alongside time-pressured Mock Interview sessions with visual countdowns and key-point checklists for metacognitive self-evaluation.
Unified Next.js full-stack architecture leveraging App Router, Server Actions, and Route Handlers, connected to a relational PostgreSQL database via Prisma ORM for type-safe query execution and authenticated with NextAuth.js.
Customized SM-2 Spaced Repetition Engine
Dynamically calculates ease factors and optimal review intervals across learning, review, and relearning card states.
Real-Time Mock Interview Simulation
Simulates interview room pressure with interactive countdown timers, urgency color shifting, and post-session performance metrics.
Checklist Metacognition & Cue Keywords
Enables users to verify their recall against structured key points and unlock gentle keyword hints when stuck.
High-Efficiency Keyboard Shortcuts & Undo
Full keyboard navigation (Space, 1/2/3, H, Ctrl+Z) with a 5-second rating undo window for lightning-fast study sessions.
Analytics Dashboard & Streak Engine
Tracks daily study consistency, retention rate graphs, card maturity breakdown, and motivational milestone badges.
Practice Out Loud Mode
Dedicated verbal articulation mode prompting candidates to speak their answers aloud before verifying against model solutions.
CHALLENGE:Adapting SM-2 to a 3-Button Evaluation System
Approach: Mapped the standard 6-grade SM-2 quality scores into an intuitive 3-button scheme (Again=1, Hard=3, Good=5) with modified interval growth formulas for Hard ratings.
CHALLENGE:Type-Safe Relational Schema for High-Frequency Progress Tracking
Approach: Modeled UserCardProgress, StudySessions, and DailyStats with Prisma ORM on PostgreSQL, adding compound indices on [userId, nextReviewDate] to ensure instant query times during card queues.
CHALLENGE:Zero-Latency UI During Rapid Card Reviews
Approach: Implemented optimistic state updates and seamless keyboard event listeners with debounce protections to keep card flipping instant.
Deployed and running live in production at https://143.198.212.224.sslip.io/ with managed PostgreSQL and automated workflows.
Successfully launched to production at 143.198.212.224.sslip.io, serving active users for technical, behavioral, and system design interview preparation.
Technologies Used
Personal Contribution
- ↳Designed the entire product vision, cognitive learning principles, and user flows
- ↳Engineered the full-stack Next.js application with Prisma ORM and PostgreSQL
- ↳Implemented the mathematical SM-2 Spaced Repetition calculation engine
- ↳Crafted the minimalist, high-contrast dark theme and responsive UI with keyboard shortcuts