Selected Projects

Background Problem

High subjectivity and inefficiency in manual cryptocurrency trading analysis and investment workflows.

Key Results & Impact

• Improved trading decision accuracy with automated short-term and long-term recommendations.
• Reduced subjective decision-making through LLM-based reasoning and quantitative analysis.

Description

Built a backend AI decision support system utilizing Python, NLP pipelines, and LLM integration to generate data-driven financial insights.

Background Problem

Manual leaflet production was highly time-consuming and heavily prone to human errors in pricing and visual layout.

Key Results & Impact

• Improved operational efficiency by over 90%, reducing production time from 12 hours to under 1 hour.
• Eliminated human errors through automated data validation and dynamic visual generation.

Description

Independently built a full-stack system using React (TypeScript) and Laravel to digitize and automate manual business workflows.

Background Problem

Required a robust, production-ready backend architecture to serve a static machine learning model to end-users.

Key Results & Impact

• Enabled real-time AI interaction by transforming a static model into a production-ready web application.
• Delivered instant predictions with low-latency inference through optimized processing pipelines.

Description

Built a scalable backend API using Python (Flask, NumPy) to seamlessly integrate the machine learning model with user interfaces.

Background Problem

Processing 40,000+ monthly operational data entries manually via Excel was inefficient and caused data inaccuracies.

Key Results & Impact

• Improved reporting efficiency by automating the processing of 40,000+ operational data entries per month.
• Reduced manual workload and increased data accuracy by replacing Excel workflows with a web-based system.

Description

Built ETL pipelines and responsive real-time dashboards using Python (Flask, Pandas), PostgreSQL, and Tailwind CSS.

Background Problem

Need for a flexible, scalable content management platform optimized for modern digital media and future AI feature integrations.

Key Results & Impact

• Enabled scalable content management and delivery by building an AI-integrated news platform.
• Improved system flexibility through a modular architecture supporting future recommendation features.

Description

Built role-based access systems, Google authentication, and traffic analytics features using React, Laravel, and PostgreSQL.

Background Problem

Need to demonstrate full-stack engineering capabilities through a fast, modern, and production-ready digital presence.

Key Results & Impact

• Developed a highly optimized, production-ready application using Server-Side Rendering (SSR).
• Enabled real-time client communication via email and WhatsApp API integrations.

Description

Developed a modern web portfolio showcasing technical projects and skills using Next.js, React, and serverless architecture.

Background Problem

Lack of engaging and accessible sign language educational tools specifically tailored for deaf children.

Key Results & Impact

• Improved accessibility to sign language education through an interactive mobile learning application.
• Enhanced engagement by delivering a user-centered design tailored for children with special needs.

Description

Designed the UI/UX for a mobile learning application aimed at supporting inclusive education and social impact.

Background Problem

Need for an accessible digital platform to assess real-time local air quality and increase public awareness of air pollution.

Key Results & Impact

• Enhanced public awareness of air pollution through integrated educational blog features.
• Delivered an intuitive interface using strict User-Centered Design (UCD) methodologies.

Description

Designed a mobile and web-based application providing real-time Air Quality Index (AQI) ratings based on live GPS data.

Background Problem

Traditional manual employee attendance and task reporting systems lacked transparency and real-time operational visibility.

Key Results & Impact

• Improved operational transparency and efficiency by digitizing task reporting systems.
• Enabled accurate real-time tracking of employees through integrated geo-tagging features.

Description

Developed a user-friendly mobile application interface using Flutter to support rapid adoption across non-technical users.

Background Problem

Academic requirement to construct a comprehensive data management web application demonstrating core web programming concepts.

Key Results & Impact

• Demonstrated robust data management capabilities with dynamic CRUD operations and state persistence.
• Successfully deployed a secure and accessible application applying OOP principles.

Description

Developed an inventory system from scratch utilizing PHP, MySQL, and JavaScript, featuring form handling, validation, and session management.

Background Problem

Participation in the competitive National Politeknik Negeri Bali Web Design Competition (PNBWDC).

Key Results & Impact

• Achieved Top 8 out of 82 participating teams in a national-level web design competition.
• Delivered a highly responsive and visually engaging user interface prioritizing accessibility.

Description

Developed a modern, responsive tourism website focusing on a strong user experience using clean HTML, CSS, and JavaScript.

Background Problem

Academic objective to practically implement Object-Oriented Programming (OOP) principles within a functional desktop application.

Key Results & Impact

• Successfully implemented complex core game mechanics, structural logic, and physical interactions.
• Executed effective team-based development and version control collaboration.

Description

Built an interactive desktop racing game utilizing Python and OOP principles to handle backend game logic.