EXPERIENCE
EDUCATION
HL: Maths,Physics,Computer Science
SL: Economics,Indonesian B,English A
SELECTED WORK
Indomahkota Portal Site
Developed a full-featured All-In-One management system for one of Indonesia’s largest contractors (outsourcing company), operating with over 500 employees across more than 10 cities.
Solved the client's previous synchronization challenges in managing inventory, invoicing, and data across multiple locations and administrative users.
Key Features:
- Multi-admin real-time inventory synchronization across all branches.
- Automated invoicing to reduce manual processing time.
- Finance, revenue and profits tracker with charts.
- Multiple roles with different permissions.
persongality.com
A fun little project I made that matches a user's spotify listening habits to personality traits.
Unfortunately this project was discontinued due to a recent rule change made by Spotify son May 15th 2025 that prevented indie developers from using their API.
Yang Bener Nih?
Founded a misinformation detection platform aimed at protecting Indonesian voters during the 2024 National Election.
Leveraging a custom-built WhatsApp chatbot, the system automatically flags false information helping users make more informed decisions.
Key Features:
- Real-time misinformation detection using a custom search engine.
- WhatsApp chatbot integration for easy user interaction.
- Multi-language support for Indonesian and English.
- Worked directly with the Ministry of Information (KOMINFO) hoax database and other fact-checking sources.
A comparative analysis of Reinforcement Learning and Depth First Search in solving mazes
Conducted an academic research project analyzing how Reinforcement Learning compares to Depth First Search (DFS) in solving mazes of increasing size and complexity.
Reinforcement Learning agents were trained using Q-learning, while DFS was applied as a baseline deterministic search algorithm. Performance metrics such as solution time, path length, and computational resources were collected and compared.
Explored the possibilities of using Reinforcement Learning to improve or prove solutions found Mathematically.
Key Features:
- Custom maze generation and simulation framework
- Q-learning based Reinforcement Learning agent
- Visualizations of agent behavior and path optimization
- Performance benchmarking across multiple maze sizes
