ABOUT_OPERATOR.txt
This dossier contains authentic, verifiable information regarding the architectural origin, design philosophy, and technical stack of the NOKY typing ecosystem.
NOKY was built by Ahmad, a Computer Science student, as a full-stack portfolio project exploring AI integration, gamified product design, and real-time systems. What started as a typing speed test grew into a complete typing platform — AI-generated content, a full progression system, and a persistent leaderboard — built end-to-end from a single developer's iterative process.
Every feature, from namespaced data partitioning to real-time custom layout states, has been iteratively refined to provide a smooth, engaging typing environment. The development timeline and system upgrades are documented transparently for peer review.
To demonstrate robust technical mastery, NOKY rejects boilerplate configurations in favor of low-latency, type-safe custom components designed for scalable web performance:
Utilizes Edge API routes, concurrent React streaming, strict asset optimization, and dynamic route handlers to achieve near-instantaneous load times.
Integrated with gemini-flash-lite-latest via secure proxy routes, employing a strict 3-second abort timeout and a zero-delay local fallback framework.
Backed by a secure remote database with row-level constraints, supporting an on-demand score submission flow to maintain clean global standings.
Implements per-player namespaced states in localStorage, retroactively computing levels, streaks, goals, and keyerror heatmaps to isolate player context.
NOKY operates with honest and transparent disclosure regarding AI-assisted content generation. We believe in providing clear visibility into how technology is utilized:
- Typing Passages: Passages for specialized sectors (e.g. Code Arena, Knowledge Quest, AI Lab, World Explorer) are generated dynamically using Google Gemini models to ensure rich, non-repetitive practice text.
- AI Coach Feedback: After each typing session, your performance metrics (WPM, precision percentage, and specific keyboard mistypes) are analyzed by the Gemini LLM proxy to generate brief, actionable coaching tips.
- Static Fallback: If Google Gemini API is offline, hits rate-limits, or times out (3-second hard limit), the platform silently falls back to local static passages and a tip loop in
lib/passages.tsto guarantee 100% uptime.
We do not present mock templates or artificial statistics. This application is supported by an open-source GitHub repository containing the complete, chronological commit logs, architectural branches, and code revisions as evidence of iterative, real-world software engineering practice: