Informatics graduate · Bandung, ID
Dimas
Data Engineer focused on reproducible data pipelines, automated batch processing, and database integrity.
Building automated ingestion workflows and analytical storage systems using Python, SQL, PostgreSQL, DuckDB, and GitHub Actions. Emphasizing clean data contracts, deterministic transformations, and verified test coverage.
Selected systems and architecture
4 systemsMECUT Radar
Automated technology news ingestion and alerting pipeline collecting from RSS feeds, Hacker News, and GitHub APIs with deduplication, keyword relevance filtering, SQLite storage, and Telegram dispatch.
Topic Modeling & Researcher Recommendation System
Controlled benchmark evaluating BERTopic with SPECTER embeddings, LDA, and NMF across 11,111 publications. BERTopic achieved the highest accuracy across all 13 recommendation metrics (NDCG@5: 0.736, MRR: 0.809), with LDA performing closely at lower compute cost.
Research Collaboration Network Analyzer
Modular desktop graphical application built with Python and PyQt6 for analyzing scientific co-authorship networks, computing network centrality metrics, and visualizing institutional research relationships.
Epidemiological Outbreak Data Pipeline
Planned batch data pipeline architecture designed to ingest public health surveillance records, enforce schema contracts via Pydantic, perform analytical SQL transformations in DuckDB, and export Parquet tables.