Muhammad Ribhan Hadiyan

01 — About

My Journey
in Data

Analytical and impact-driven Data Science & Modelling Specialist at BFI Finance with a strong Mathematics background (GPA 3.88/4.00). Expert in the full ML lifecycle — from engineering complex features in Trino/ODPS to deploying predictive models for mortgage and retail assets.

Proven track record in automating Daily Model Monitoring systems and real-time data infrastructure alerts via G-Chat integration. Certified TensorFlow Developer skilled at bridging advanced research (Fuzzy RBM, Deep Learning) with scalable financial solutions that optimize collection strategies and operational efficiency.

My academic journey at Universitas Padjadjaran forged a rigorous analytical mindset, further deepened through research in Operations Research and Robust Optimization, culminating in a thesis achieving 92% accuracy in malaria cell image classification.

3.88 GPA / 4.00

4+ Publications

900+ Hours ML Training

Technical Skills

Python (Pandas, NumPy, Scikit-learn)

SQL — Trino & ODPS

R & LaTeX

Deep Learning — TensorFlow / Keras

MLOps & Model Deployment

Streamlit & Apache Superset

Feature Engineering & Optimization

SHAP — Explainable AI

02 — Experience

Work
History

Nov 2024 — Present 01

Asset Management Data Science & Modelling Specialist

PT BFI Finance Indonesia, Tbk

Tangerang, Banten, Indonesia

Led end-to-end development of 2 Mortgage Predictive Models classifying 1,500+ monthly contracts — managing the full lifecycle from migrating 200+ features from ODPS to Trino through to management approval.
Orchestrated complex Model Deployment & QA, executing 7 UAT scenarios and 8 rollback contingency scenarios for retail model deployment (Car/MCY) to ensure system stability.
Developed a dual-layer Model Monitoring System: a Streamlit app for Data Science deep-dives and Apache Superset dashboards for business stakeholders, backed by a Trino SQL Datamart.
Re-engineered "Is Success" Model Tracking — transformed feedback loop from a 15-day lag to a Daily Monitoring System, enabling real-time performance evaluation.
Automated infrastructure health checks via G-Chat Integration, monitoring 5 critical tables daily with instant alerts for data pipeline issues or delays.

Feb 2023 — Feb 2026 02

Research & Teaching Assistant

Universitas Padjadjaran · with Prof. Diah Chaerani

Jatinangor, West Java, Indonesia

Conducted bibliometric analyses on 5+ research topics (robust optimization, machine learning, green economy) using R-Biblioshiny & VOSviewer, analyzing 200+ articles to support 3+ journal publications.
Formatted and revised 15+ manuscripts for international journals (IAENG, Elsevier, Springer) using LaTeX, ensuring compliance with complex templates and incorporating reviewer feedback.
Modernized Optimization & Nonlinear Programming curriculum — migrated from Maple/MATLAB to Python, creating OBE-based lesson plans and assessments for 60+ students.
Contributed to a textbook on "Optimization with Python" including visualization code, indexing, and glossary.

Jan 2022 — Dec 2023 03

Laboratory Assistant

Asisten Laboratorium Matematika Unpad

Jatinangor, Jawa Barat, Indonesia

03 — Work

Featured
Projects

Project / 001

Predictive E-wallet Fraud

A real-time interactive dashboard for detecting and analyzing fraudulent e-wallet transactions using machine learning and explainable AI (SHAP).

Python Dash SHAP

View on GitHub

Project / 002

Sentiment Analysis — Megathrust

Processed 10K+ tweets using NLP pipelines to classify public sentiment on megathrust earthquakes using the Indonesia-BERT model.

Python BERT NLP

View on GitHub

Project / 003

LiFit — BMI Detection App

Capstone mobile app for BMI detection and health tracking with personalized recommendations powered by machine learning.

Kotlin TensorFlow Firebase

View on GitHub

Project / 004

Stock Prediction — RNN & LSTM

Built RNN & LSTM models for stock market analysis. RNN achieved MAE of 0.0249, demonstrating strong time-series forecasting capability.

Python RNN LSTM TensorFlow

04 — Research

Publications
& Research

2025 01

Fruit Ripeness Classification Using CNN & Bibliometric Analysis

IAENG International Journal of Computer Science

CNN · Computer Vision · Bibliometric Analysis

2025 02

Benders Decomposition on Multi-objective Integer Adjustable Robust Counterpart Optimization with Polyhedral Uncertainty Set

Engineering Letters

Operations Research · Robust Optimization · Integer Programming

2025 03

Lontar Printing Optimization for Ancient Sundanese Manuscript Preservation

Engineering Letters

Optimization · Cultural Heritage · Operations Research

2024 04

Fuzzy RBM Feature Extraction on Fashion-MNIST

Jurnal Sistem Informasi Bisnis

Deep Learning · Fuzzy Logic · Feature Extraction

05 — Education

Academic
Background

Aug 2020 — Jun 2024 01

Bachelor of Mathematics — GPA 3.88 / 4.00

Universitas Padjadjaran

Jatinangor, Indonesia

Thesis: Implementation of Mixed Accelerated Learning based on Fuzzy Restricted Boltzmann Machines and SVM for Malaria Cell Image Classification — 92% Accuracy.
Achievement: First Runner-Up, Outstanding Mathematics Student Award 2023.

Feb 2023 — Jul 2023 02

Machine Learning Specialization

Bangkit Academy · by Google, GoTo & Traveloka

Bandung, Indonesia

Completed 900+ hours of rigorous training in Machine Learning, Deep Learning, and Cloud Deployment.
Achieved Google TensorFlow Developer Certification on the first attempt.

Muhammad
Ribhan Hadiyan

My Journey
in Data