Back
_
1 Teams Member
A high-concurrency mobile portal for enterprise workforces featuring Redis-buffered geofenced attendance, offline-ready document management, and instant access to immutable payroll snapshots for 500k+ users.
Project Challenge:
Super-HR was engineered to eliminate manual data-entry and payroll errors by replacing traditional human processing with deterministic automation waterfalls. The project focused on a decentralized management model where logic is pushed to the 'Edge' (Hiring Managers). Key challenges included building a multi-tenant payroll engine that handles mid-month proration, tiered Indonesian labor law overtime (PP 35/2021), and a mathematical delta engine to detect unexcused absences across millions of records without daily cron jobs.
Steps:
- Auto-screen and score high-volume CVs using Gemini AI to find top-tier talent.
- Empower Managers to approve onboardings, attendance anomalies, and wage-waivers.
- Execute the EOM payroll waterfall with a 100% eradication of compound penalty errors.
- Manage regional/national holiday sets with dynamic regional ISO code matching.
- Export bank-ready bulk transfer CSVs generated from immutable ledger snapshots.
Tech / Framework:
- Next.js (T3 Stack)
- Prisma + PostgreSQL (Deep Indexing)
- Google Gemini AI (Candidate Scoring)
- Tailwind CSS (Admin Dashboard)
- Zod (Strict Financial Validation)
Achievement:
I architected the core Payroll & Automation Engine, implementing a three-pass in-memory processing strategy for 500k users. I designed the database schema with materialized paths for organization tree scaling and optimized string-indexed date constraints for $O(log n)$ query performance. I also built the AI integration layer for automated recruitment and developed the deduction hierarchy logic to ensure 100% legal compliance in automated pay-cuts.
Nur Qalik 2026 © All right reserved.