RETNO SETIAWAN

Line Manager Control Tower

Control Tower

Operating Point Coordinator

District Coordinator

Coordinator Operational

First Mile Data Entry

First Mile Driver

First Mile Sorter

Logistics Operations Management

Fleet Operations, Control Tower Functions, First Mile Logistics, Multi-branch Network Management, SLA Performance Improvement, Operational Reliability, High-Volume Logistics, Operational Discipline, KPI-driven Monitoring, Root Cause Analysis, Operational Risk Control, Process Optimization

Data Analysis & Operational Reporting

Operational Data Analysis, Fleet Utilization Analysis, KPI Analysis, SLA Monitoring, Delivery Metrics Analysis, Performance Reporting, Real-time Monitoring, Data Integrity, Fraud Detection

Fleet & Transportation Management

GPS Monitoring, Route Accuracy, Operational Visibility, Unit Performance Analysis, Idle Time Analysis, Route Deviation Analysis, Driver Behavior Analysis, Fleet Utilization Tracking, Transport Management System (TMS)

Process Improvement & Cost Efficiency

Process Improvement Initiatives, Operational Cost Leakage Reduction, Fraud Prevention, Stricter Controls Implementation, Efficiency Enhancement, Automation Development, Manual Audit Reduction, Compliance Monitoring

Team Leadership & People Development

Cross-functional Team Leadership, Team Supervision, Performance Development, Coaching & Training, Workload Distribution, Team Scheduling, Staff Performance Management, Conflict Resolution

Automation & Technology

AI-powered Automation, Attendance Validation System, Computer Vision Analysis, Fraud Detection Logic, Node.js, Playwright, TensorFlow.js, Face-API.js, SheetJS, Telegram Bot API

I developed an AI-powered automation system to enhance operational control and reduce manual audit processes, demonstrating strong problem-solving capability and a technology-driven mindset in logistics operations. This system was built to address recurring fraud cases in driver attendance, where primary drivers falsely reported operating with a second (freelance) driver who was not actually present. In several cases, drivers manipulated attendance by logging in with multiple devices and submitting invalid check-in photos (e.g., non-selfie images, road photos, or obscured faces) to simulate the presence of a second driver. This behavior led to multiple operational risks, including misuse of driver allowances (allocated for two drivers but used by one), increased fatigue due to single-driver operations, and ultimately delivery delays impacting SLA performance and return schedules. To mitigate these issues, I developed an end-to-end automated attendance validation system capable of detecting non-compliant check-ins through computer vision analysis and fraud detection logic. The system identifies anomalies such as missing or unclear facial presence, multiple individuals, and potential replay attacks (duplicate photo reuse), enabling faster and more reliable validation. The solution also automates data extraction, validation, and reporting, significantly reducing manual audit time while improving operational visibility and compliance control across the network. Tech Stack: Node.js, Playwright, TensorFlow.js, Face-API.js, SheetJS, Telegram Bot API.