I'm a final-year Information Systems undergraduate at Universitas Indonesia interested in data analytics, data engineering, and applied machine learning. I build portfolio and university projects with an emphasis on data quality, reproducible workflows, clear documentation, and usable reporting.

• Main tools: Python, SQL, BigQuery, pandas, Excel, Looker Studio, and Tableau
• Also worked with PostgreSQL, Django REST Framework, Java, Spring Boot, Vue, Git, and Docker
• Currently completing a reproducible Indonesian food visible-ingredient benchmark
• Open to Data Analyst, Data Engineering, Data Science/ML, and Software Engineering internship opportunities

Repository · Interactive dashboard

An end-to-end retail analytics, data warehousing, and forecasting project built from 3,000,888 store-family-day records across 54 stores and 33 product families.

• Built raw, core, and analytical mart layers in BigQuery, documenting table grain, keys, join rules, and missing-data treatment
• Developed Python and SQL workflows for auditing, transformation, feature preparation, and validation
• Produced 28,512 forecast records and downstream Looker Studio and Excel reporting layers
• Included rolling-origin model validation and 17 reconciliation and data-quality checks

Technologies: Python, SQL, BigQuery, pandas, LightGBM, Microsoft Excel, Looker Studio

Repository · Interactive project site

A human-annotated NLP study examining how sentiment expressed in review text relates to Steam's binary recommendation.

• Analyzed 262 de-identified English reviews labeled from review text alone
• Independently double-annotated 80 reviews and adjudicated all annotation disagreements
• Used duplicate-aware data partitions, repeated grouped validation, and a locked evaluation set
• Packaged the analytical workflow into reusable Python modules with automated tests and GitHub Actions
• Documented sampling, annotation, model, and generalization limitations

Technologies: Python, pandas, scikit-learn, TF-IDF, Logistic Regression, statistical analysis, GitHub Actions

Repository

An ongoing benchmark comparing CNN and vision-language model predictions under the same visible-ingredient ontology and evaluation protocol.

• Designed a 43-label ontology for visually supported food components
• Developed reproducible screening, annotation, adjudication, training, and evaluation workflows
• Added validation scripts, integrity tests, experiment configuration, and responsible-use documentation
• Final benchmark results remain unpublished while annotation and adjudication are incomplete

Technologies: Python, PyTorch, Hugging Face Transformers, computer vision, vision-language models

A five-person team project developed to digitize correspondence management, approvals, status tracking, and document archiving for a school.

• Served as Lead Programmer and defined the system architecture and development environment
• Coordinated frontend–backend integration and supported API and UAT validation across three sprints
• Worked with Django REST Framework, Vue, PostgreSQL, Supabase, Cloudflare R2, Railway, and Vercel

Technologies: Python, Django REST Framework, Vue 3, TypeScript, PostgreSQL, REST APIs

• Data and analytics: Python, SQL, BigQuery, pandas, NumPy, Microsoft Excel, Looker Studio, Tableau
• Machine learning: scikit-learn, LightGBM, PyTorch, NLP, forecasting, computer vision
• Backend and databases: Django REST Framework, Spring Boot, PostgreSQL, MySQL, REST APIs
• Development tools: Git, Docker, Postman, GitHub Actions

LinkedIn · Email