Project archive
Full collection of 12 data science, analytics engineering, automation, and BI projects.
Streamlit + Plotly dashboard scoring 954 BEI-listed companies across 11 strength and growth metrics. Custom formula builder, sector comparison, Parquet caching cuts query time 10×.
Scraped 1,608 articles from 11 Indonesian outlets into BERT/LLM/NLP-ready datasets. Shipped sentiment analysis, topic modeling, spike detection, and executive summaries.
Selenium network-interception workflow extracts 958 IDX stocks, 45 indices, and 11 sectors, then auto-generates publication-ready Indonesian market reports on a recurring schedule.
Consolidated data engineering suite: ETL framework parsing 20+ Indonesian economic indicators from BPS, Bank Indonesia, Kemenkeu, and DJPPR archives (2003–2025); reusable scraping stack processing 30,000+ BKI ship records and 15,000+ school records across 39 provinces with retry logic, rate limiting, and checkpoint/resume; automated recurring report feeds for SPIP credit card, GAIKINDO automotive, and World Bank manufacturing indicators from multi-sheet Excel and PDF sources.
Looker Studio dashboard modeling 3 World Bank manufacturing indicators across 265 countries and 65 years. Country-year long format, consistent metric definitions, and multi-dimensional reporting for trend and contribution analysis.
Looker Studio dashboard for Indonesian disaster statistics — kejadian bencana by province and disaster type. Multi-year trend and contribution analysis for editorial and policy stakeholders.
News intelligence workflow for PLTN and renewable energy research: keyword scraping, RSS collection, deduplication, source tracking, and structured analysis notebooks for long-form reporting.
Bangkit 2024 capstone. Image-based nutrition analysis: dataset acquisition and preprocessing, transfer learning experiments, H5 model export, lightweight Flask inference API.
Deep-learning IDS for a government Super App — DNN compared with Multinomial Logistic Regression, Naive Bayes, and AdaBoost. 95% detection across DoS, Probe, R2L, and U2R categories. 3rd runner-up at TSDN 2023.
Interactive Tableau dashboard analyzing Adidas sales across retailers, products, and regions. Dynamic KPIs, contribution analysis, country/product filters, and stakeholder-ready storytelling with YoY trends and profitability metrics.
Executive-level Tableau dashboard for Classic Models — global sales performance, product-line profitability, office rankings, and customer segmentation. Dynamic parameter controls and drill-down hierarchy for strategic decision-making.
Tableau dashboard tracking Coffee Chain profit drivers across product categories, locations, and time. Margin analysis, market-basket insights, and interactive what-if scenarios for operational and inventory optimization.
