Hey, I'm Farrel Caesarian
I dive in |
Informatics Engineering student at State Polytechnic of Malang. Building scalable backend architectures, developing AI-powered features, curating high-precision data pipelines, and structuring elegant interfaces.
<IT/>
Diving Into The Tech Matrix
M. Farrel Caesarian
Informatics Student & Developer
My Mission
I'm an Informatics Engineering student at the State Polytechnic of Malang. My technological interests span Full-stack Web Development, Data Science, Computer Vision, and Artificial Intelligence.
I specialize in engineering reliable, highly scalable backend software alongside accurate machine learning pipelines, building high-impact products that solve concrete, real-world problems.
Core Competence & Technical Stack
Designing seamless user interfaces that offer high visual hierarchy, clean responsive grid cards, and fluid interactive layers.
Writing highly modular service logic, structuring secure relational databases, and microservices server nodes.
Orchestrating cross-functional team collaborations, presenting insights, and aligning resources dynamically.
Performing deep exploratory data analysis, building visualizations, and training computer vision or classification models using state-of-the-art architectures and hyperparameters.
Modern Career Roadmap
Software Engineer Intern
PT Kosmetika Global Indonesia- Developed a cosmetics raw material weighing system using MQTT & RS232, enhancing speed and accuracy by **20%**.
- Designed a digital Certificate of Analysis (CoA) module, achieving a **70%** manual error reduction with real-time data tracking.
- Implemented real-time computer vision models to scan expiration and batch details, mitigating manual inspection overhead.
- Demonstrated Kosmesia’s AI Assistant (KIVA) to enterprise clients, improving consultation speed.
Project Manager & ML Engineer
State Polytechnic of Malang- Managed a cross-functional team of 5 to construct a facial expression classifier model from dataset collection to deployment.
- Trained a Support Vector Machine (SVM) reaching **92% recognition accuracy**, optimized for real-time video feeds.
- Organized resource milestones to successfully deliver target models within a strict 1-month timeline.
Data Analyst Lead
DQLab (Business Decision Research)- Analyzed complex retail customer datasets to identify behavioral purchase correlations.
- Configured predictive regression structures and visual analytics using Pandas and Matplotlib.
Project Manager & Full-Stack Developer
State Polytechnic of Malang- Coordinated a 5-member team to deliver web frameworks using Laravel and MySQL.
- Managed SDLC schedules, presented progress updates, and served as primary liaison.
Front-End Web Developer
State Polytechnic of Malang- Evaluated user interaction data to design clean, high-performance responsive web pages.
- Optimized rendering times and validated layouts for complete cross-platform stability.
Engineered Projects & AI Models
HealthGrade AI Nutrition Advisor
HealthGrade scans packaging barcodes to display Nutri-Grade (A–D) health indexes, featuring AI OCR scanning and advice chatboxes (SDG 3 & 12 focused).
Jakarta's Air Pollution Classifier
Predicted daily air quality in Jakarta with 93% accuracy using ML models. Utilized IBM Granite AI to summarize local logs and classify causes into target groups.
HashConv Real-Time Chat
A full-stack, secure real-time messaging portal incorporating a Laravel API framework coupled with a highly responsive, fluid React.js client interface.
Chemulation VR Simulation
A fully interactive VR chemistry lab allowing students to safely trigger reactions and visual atomic structures inside high-precision virtual environments.
Customer Churn Prediction
Developed a machine learning model predicting customer churn rates for video streaming services, achieving 75% prediction accuracy to support retention actions.
Food Image Classification
Food classification competition using pre-trained EfficientNetV2B1 networks to achieve 81% accuracy across 15 diverse food item classes.
Coral Species Classification
Classified unique underwater coral species using Custom Convolutional Neural Network models, securing a 78% accuracy rate on marine data.
Temperature Prediction Model
Built weather prediction models supporting farming metrics, incorporating Optuna hyperparameter sweeps and MICE missing value cleaning (Rank 11/88).
Sales Insights BI Dashboard
Created comprehensive, high-fidelity BI charts mapping customer distributions and sales trends, helping stakeholders make quick data decisions.
Internify Internship Finder
A dedicated web portal assisting college students to discover, filter, and easily apply for verified corporate internships.
Sibeta Thesis Manager
A fast, native PHP web system designed to orchestrate thesis proposals, examiner scheduling, and PDF submissions in university departments.
Face Recognition OpenCV
A computer vision tool built to process and identify faces inside static and live camera feeds using OpenCV classifiers.
Establish Core Connection
Contact Channels
Get in touch or follow my developer nodes across the net. Ready for collaborations and direct interfaces.