I built an AI-first Fund Analysis System that enables natural-language financial intelligence through a Retrieval-Augmented Generation (RAG) architecture. The platform automatically ingests and structures financial PDF documents, computes institutional metrics such as IRR, DPI, and PIC, and exposes an AI conversational interface powered by LangChain, Gemini LLM, and pgvector-based semantic retrieval. The system is architected with modular FastAPI microservices, fully containerized using Docker for scalable deployment, and integrated with a Next.js frontend to deliver a seamless user experience while being engineered with production-level reliability and extensibility in mind.

This project is an e-commerce platform enhanced with artificial intelligence (AI) to simplify product management. The system is designed so that store admins no longer need to manually input product details. By simply uploading a product image, the AI automatically predicts the product’s name, description, and estimated price. If the prediction is not fully accurate, admins still have the flexibility to edit the details before publishing. The project leverages modern technologies, including n8n for workflow automation, Supabase for backend and database management, and Gemini as the AI model for generating product names and descriptions.

This traffic analysis system utilizes a YOLOv8 model trained to detect cars and motorcycles, which is then converted to TensorFlow.js to run directly in the browser. Based on the detection, a PCU (Passenger Car Unit) algorithm calculates the green light duration proportionally, ensuring the most congested lanes get more time to reduce traffic jams. This project is still in development, but it provides a clear picture of AI's potential to optimize urban traffic.

Built an AI-powered waste classification app using Vision Transformer (ViT), trained on 11,050 images across 3 categories (organic, non-organic, hazardous). This solution helps simplify waste sorting at the source, reducing contamination, improving recycling efficiency, and supporting a cleaner, more sustainable planet.

I led the UI/UX team in designing an information system application for Student Activity Units (UKM). My main responsibility was to translate user needs into intuitive interface designs while ensuring a consistent user experience across the platform. The application features a streamlined facility borrowing flow in the logistics section, program management for proposals and progress tracking, as well as UKM monitoring through KPI dashboards, leaderboards, and media documentation. It also provides membership management with full CRUD functionality, and open recruitment. Lastly, the collaboration section supports media partnerships, sponsorship tracking, and a job board for opportunities.

Developed a responsive and interactive company profile website to enhance user experience, effectively showcase the company's services, and drive business growth.

I was given the opportunity by PT. Bogart Inti Perkasa to design a desktop-based application with features for recording income, expenses, tracking receivables and accounts receivable, and managing correspondence complete with document tracking. I also built features to support the smooth operation of the company.

Recognizing the critical need for efficient emergency response, we developed Aegis, a cutting-edge emergency service application designed for rapid alerts and seamless integration with smart city systems.

The system centralizes citizens' personal data, including names, addresses, phone numbers, and other essential information, in a secure and easily accessible database. A map-based visualization feature enhances administrative efficiency by displaying each resident's home location, making it easier to identify key areas and streamline local governance. Additionally, a document management module allows for the creation, storage, and organization of official village correspondence, ensuring smooth administrative processes. To promote financial transparency and better budget management, the system also includes features for recording village income and expenses in a structured and organized manner.

This website serves as a comprehensive platform for displaying real-time data on COVID-19 cases from various countries around the world. It provides users with up-to-date information, enabling them to monitor global trends, and analyze the spread of the virus.