I am a Lead Backend Engineer specializing in Distributed Systems, Human Capital Management (HCM) Architectures, and Industrial IoT. With a background in Electronic Engineering.

My focus is on building resilient, high-concurrency systems that solve critical business problems—from automating complex payroll logic for thousands of employees to orchestrating industrial machinery via Python.

• Core Stack: Python, Django/DRF, AWS, Docker, PostgreSQL, Redis, Celery, Linux.
• Key Achievement: Architected the system that enabled SMETA Audit Compliance for two consecutive years, securing international export licenses.
• Location: Maracaibo, Venezuela (Open to Remote Roles globally).

⚠️ Note: This is a portfolio showcase repository for a proprietary system engineered for Grupo Confremarca. The full source code is private. This repository highlights the architectural decisions, patterns, and modules designed to solve complex Human Capital Management (HCM) challenges at scale.

Confremarca Enterprise ERP is a high-availability, distributed Human Capital Management (HCM) system engineered to serve thousands of employees across multiple industrial plants. It replaces legacy manual processes with an automated platform handling everything from biometric attendance to complex payroll calculations.

• Role: Sole Architect & Lead Engineer
• Stack: Python, Django, Celery, Redis, Docker, PostgreSQL, AWS
• Status: Production (Active since Feb 2024)

The system is designed as a modular monolith using Django as the core, decoupled from heavy processing tasks via Celery and Redis. It integrates with physical hardware (Biometric IoT) and legacy ERPs.

High-Level Data Flow

1. Ingestion Layer: Biometric devices push logs to an intermediary buffer.
2. Async Pipeline: Celery workers pick up logs, apply "thundering herd" protection (Mutex locks), and normalize data.
3. Core Logic: Django models process standardized events into "Attendance," "Shift," or "Absence" records.
4. Sync Layer: Validated data is pushed to the legacy ERP for accounting consistency.

• Challenge: At shift changes (e.g., 7:00 AM), thousands of employees clock in simultaneously. The legacy system would crash or create duplicate records due to race conditions during sync.
• Solution: Implemented a Distributed Lock (Mutex) pattern using Redis.
  • A dedicated Celery task acquires a non-blocking lock redis.set(..., nx=True).
  • If a sync cycle is already running, subsequent triggers "fail fast" and wait for the next cycle.
• Result: Zero duplicate records and 100% data integrity for compliance audits.

• Challenge: Standard libraries couldn't handle industrial shifts like "2 Days, 2 Nights, 2 Rest" with dynamic overtime calculation spanning across midnight.
• Solution: Engineered a custom Shift Projection Engine.
  • Mathematical models project the shift pattern indefinitely into the future.
  • Used Strategy Pattern to apply different overtime rules based on shift type (Diurnal/Nocturnal/Mixed).

• Challenge: International exports required a tamper-proof record of working hours and breaks.
• Solution: Built an immutable "Audit Log" module.
  • Every modification to an attendance record creates a shadowed log entry.
  • Digital Signature workflow ensures employees legally validate their overtime hours.
• Impact: The system was the primary evidence used to pass the SMETA audit for 2 consecutive years.

• Attendance & Shifts: Real-time tracking of entry/exit with geolocation and biometric verification.
• Payroll Engine: Automatic calculation of variable pay, bonuses, and statutory deductions.
• Digital Signature: Legal workflow for document acceptance.
• Industrial Dining: Logistics module for cafeteria management and food benefit distribution.

Adolfo Jimenez | Lead Python Backend Engineer

• 🌐 adolfoj.com
• 👔 LinkedIn
• 📧 adolfoj@protonmail.com