I’m a Full-Stack Engineer with strong business analysis and system design experience, focused on translating complex business requirements into scalable, production-ready systems that deliver real operational impact.
I hold a Bachelor of Computer Science (S.Kom.) in Information Systems from Atma Jaya Yogyakarta University, with graduation pending in May 2026. My background allows me to bridge technical implementation with business needs, ensuring solutions are both technically sound and context-aware.
I have hands-on experience applying machine learning in real systems, including hybrid recommendation engines, computer vision models achieving 91%+ F1-score, and automated ETL pipelines deployed to production. I treat machine learning as a system component that is designed, integrated, and maintained, not as a standalone experiment.
Previously, I completed a 5-month AI/IT internship at PT Salam Pacific Indonesia Lines, where I worked on refactoring enterprise systems for real-time container operations, migrating video streaming infrastructure for AI-based damage detection, and delivering production frontends using Next.js, Vue.js, Golang, and Python.
I’m currently seeking Full-Stack or ML-focused roles where I can contribute as an engineer who understands both the business context and the technical depth required to build scalable, reliable systems.



"In Every Line of Code, Plant a Seed of Knowledge"
— Mark Alvin's Personal PhilosophyShowing 10 of 10 projects
Web-based hospital management system for diabetes patient care at RS Panti Nugroho. Implemented role-based access control for 9 distinct medical and administrative user roles, including separate roles for outpatient and inpatient nurses.
View WebWeb application translating Indonesian Sign Language (BISINDO/SIBI) to text. React.js frontend with real-time MediaPipe camera integration. Flask REST API with dual ML models achieving 91%+ F1-score.
View WebComprehensive Point of Sale system for coffee shop operations. Real-time inventory management with automated stock alerts. Reduced order processing time by 40% and eliminated manual inventory errors.
View WebWeb-based POS system for Japanese BBQ All You Can Eat restaurant. Next.js with responsive design and secure authentication system. Collaborative development with 3 full-stack developers.
View WebHybrid system combining Content-Based Filtering and Neural Collaborative Filtering. Rated "Excellent" in Dicoding Machine Learning final project assessment. Advanced feature engineering with book metadata and user-item interactions.
View GithubAutomated web scraping using BeautifulSoup and Requests libraries. Multi-format storage: CSV, PostgreSQL, Google Sheets integration. Comprehensive unit testing suite ensuring pipeline reliability.
View GithubML classification pipeline with SMOTE handling for dataset imbalance. Comparative model evaluation achieving high predictive accuracy. Comprehensive EDA revealing actionable churn patterns.
View GithubComputer vision solution using pre-trained InceptionV3 architecture. Multi-format deployment: SavedModel, TensorFlow Lite, TensorFlow.js. Classification of 6 natural scene categories with data augmentation.
View GithubEnd-to-end ML pipeline integrating clustering and classification methodologies. K-Means clustering with silhouette score optimization (≥ 0.70). Achieved F1-score and accuracy ≥ 99.85% on test data.
View GithubLive interactive dashboard revealing customer behavior patterns. Seasonal and weather impact analysis on rental demand. Business recommendations for operational optimization.
GitHub RepoLive DashboardFeel free to reach out to me through the following platforms: