Open to work · Freelance available
Bugi Sulistiyo
Data Scientist / Data Analyst / ML Engineer
Embracing challenges with curiosity and determination. Continuously learning, I strive to deliver solutions that are thoughtful and effective. I enjoy transforming ideas into results that make a real difference.
- Python
- Pandas
- TensorFlow
- Keras
- Scikit-Learn
- OpenCV
- SQL
- BigQuery
- Looker Studio
- Metabase
- Matplotlib
- Git
Experience
Where I have worked and what I built there.
Oct 2025 - Sept 2026
Digital Analyst
PT Cahaya Pengajaran Abadi · Full-time
Balikpapan, East Kalimantan, Indonesia · On-site
- Pioneered the company's data capability by independently designing and implementing a full-scale data warehouse, covering planning, architecture, modelling, and end-to-end pipelines.
- Delivered actionable insights through dashboards and reports on app usage behaviour and mining production performance, supporting data-driven decisions in fleet management operations.
- Partnered with cross-functional teams to align business needs with data solutions, enabling better outcomes for both internal and external stakeholders.
Apr 2025 - Jul 2025
Data Scientist
MINERGO SYSTEMS · Contract
Balikpapan, East Kalimantan, Indonesia · On-site
- Designed and built a full-scale data warehouse from scratch across 2 source systems, covering architecture, data modelling, and end-to-end ETL pipelines from raw to DWH to data mart layers.
- Engineered Apache Airflow pipelines for hundreds of PostgreSQL tables handling terabyte-scale data, with the largest table generating 5-8 GB/day, ensuring reliable incremental loads and data integrity.
- Delivered interactive Metabase dashboards on app usage and mining production metrics, directly supporting operational decision-making for internal and external stakeholders.
Feb 2025 - Jul 2025
Machine Learning Mentor
Coding Camp powered by DBS Foundation · Part-time
Indonesia · Remote
- Supported students in completing capstone projects while fostering technical, analytical, and professional skill development in a collaborative environment.
Sept 2023 - Dec 2023
Data Engineer
PT Bank Aladin Syariah Tbk · Internship
Jakarta, Indonesia · Hybrid
- Built the company's first real-time BigQuery monitoring dashboard using Looker Studio and Slack alerting system, surfacing inefficient query traffic patterns and improving cost efficiency.
- Improved peak-hour traffic visibility, enabling the team to proactively optimise query scheduling and reduce failure rates.
Feb 2023 - Jan 2024
Machine Learning Path Mentor
Bangkit Academy led by Google, Tokopedia, Gojek, & Traveloka · Part-time
Indonesia · Remote
- Mentored students in TensorFlow, deep learning, and computer vision, providing technical guidance, code reviews, and best practices for model development.
- Led discussions and workshops to bridge theory with real-world applications, helping students develop industry-relevant capstone projects.
- Fostered a collaborative and engaging learning environment, encouraging critical thinking and problem-solving to support student success.
Project
Segmentation, classification, and sentiment models built on real datasets, each linked to the work itself.
12 Dec 2024
Semantic Segmentation of Glaucoma Disease. Using CDR Value to Identify Glaucoma on Fundus Image.
A research project utilising retinal fundus images collected at a hospital in Samarinda. A custom U-Net model with augmentation (CLAHE) was developed to segment the optic cup and disc, enabling the calculation of the cup-to-disc ratio (CDR) as the basis for glaucoma detection. Weighted labels were applied to handle class imbalance during training. The method achieved an AUC of 85,29% and provided a more interpretable decision-making process compared to direct image classification. This approach demonstrated potential as a reliable screening tool to assist doctors in glaucoma assessment.
08 Jun 2024
Comparison of EfficientNet and MobileNet on Glaucoma Detection Using Fundus Image
A research study comparing MobileNetV2 and EfficientNet architectures for glaucoma classification based on retinal fundus images. The project involved preparing medical dataset, applying image enhancement and augmentation techniques (resizing, rescaling, flipping, brightness, and CLAHE), and evaluting perfomrance across multiple scenarios. Model were fine-tuned using transfer learning and assessed with metrics including AUC, F1-score, accuracy, and precision. Results showed that MobileNetV2 achieved the best overall performance, highlighting the importance of data preprocessing and fine-tuning over model complexity.
08 Jun 2022
CatPedigree - Cat Breed Classification and Pedigree Prediction
A capstone project developed at Bankgit Academy, designed as a cat dating app with machine learning features for breed classification and pedigree prediction. The ML system was co-developed, covering data preparation, model training, and evaluation. Transfer learning Techniques were applied to classify cat breeds with accuracy above 80%. Pedigree prediction was implemented using a custom dataset generated with the CatBreeds calculator and modelled with a lightweight ANN. Both models were converted to TensorFlow Lite and integrated into the mobile app prototype as part of the MVP feature set.
03 Feb 2025
Predictive Analytics of User Sentiment on the MitraDarat Application
Analysing public sentiment toward the MitraDarat app, an official platform by the Ministry of Transportation for real-time city bus tracking. user reviews were collected through the Google Play API, resulting in a dataset of over 1.900 entries. The data underwent extensive preprocessing, including text cleaning, stopword removal, stemming, slang normalisation, and lexicon-based sentiment labelling. Multiple machine learning and deep learning models were evaluated, with LSTM achieving the best accuracy at 84,96%. The analysis revealed that while most users expressed positive sentiment, a significant portion highlighted recurring issues that require attention. These insights can help guide improvements to enhance user satisfaction and the reliability of the app's core features.
04 Nov 2024
Image Classification for Trash Sorting
Using CNN to classify images of waste into five categories (cardboard, glass, metal, plastic, and trash). The model was trained using the TrashNet dataset and achieved an accuracy of over 70%. This project explored the challenges of working with relatively small datasets, highlighting the importance of augmentation and careful label selection for improving classification performance in real-world applications such as recycling or waste management.
09 Oct 2024
Sentiment Analysis of User Reviews on the BRImo Application
Applied machine learning to analyse thousands of customer reviews from the BRIMO mobile banking app. The reviews, written in both Indonesian and English, are processed to remove noise and slang before being classified into positive, neutral, or negative sentiments. The analysis reveals overall user satisfaction and highlights areas where the app can improve, with the best model achieving nearly 90% accuracy.
Education
Formal background in informatics.
Aug 2019 - Jun 2024
Universitas Mulawarman
Bachelor's degree, Informatics
Activities and societies: KSR Universitas Mulawarman, Association of Computer Science Students, Association of Informatics, BEM FKTI
Sept 2021 - Dec 2021
Institut Teknologi Indonesia
Exchange Students, Informatics
I got the opportunity to get the learning experience of another campus from Kampus Merdeka Programs. Besides studying academically, I also take part in Modul Nusantara which introduces the culture and history of Tangerang Selatan. Class taken: - Digital Forensic - Machine Learning - Teknologi Multimedia - Visual Programming
Area of Interest
Medical imaging, language, and the analysis that turns both into decisions.
Analysis
I explore and analyze diverse datasets—tabular, text, or image—using statistics and insights-driven methods. My work uncovers patterns and relationships that inform effective decisions and modeling strategies.
Deep Learning
Specialized in image classification, I build models from scratch, leverage transfer learning, and optimize architectures for each case. I handle everything from data augmentation to deployment-ready solutions.
Visualization
I create dashboards that transform complex data into clear, actionable insights. My visualizations are intuitive, automatically update with data, and make information easy to understand.
Medical & Health Analytics
Focused on eye disease research, I analyze clinical and patient datasets to extract meaningful insights. My work combines domain knowledge with data-driven methods for impactful healthcare solutions.
Research & Experimentation
I tackle complex problems through applied research and hands-on experimentation. From algorithm development to novel implementations, I explore innovative solutions to real-world challenges.
Computer Vision
Experienced in image classification, object detection, semantic segmentation, and feature extraction. I manage the full pipeline—from data annotation and augmentation to image processing and algorithm development.
Automation
I streamline workflows and data pipelines, turning repetitive tasks into efficient, automated processes. From spreadsheets to dashboards, I ensure solutions save time while delivering reliable results.
Publication
Research written up in full, including a book on computer-aided glaucoma detection.