Vrica Gede Penggalih

A interpreted programming language written in Go with syntax focused on the Indonesian Language. Features its own lexer, parser, evaluator, and detailed line-number error reporting.

• Data types: string, integer, boolean, and arrays
• High-order and first-class functions support
• Conditional statements and rich expression evaluation
• Designed language grammar and execution model from scratch
• Includes online web playground & CLI execution engine
• ~3000 lines of handwritten code (didn't use AI)

An enterprise Document Management System created for client PT Aino Indonesia as part of my University Capstone Project. I serve as the only developer on the team.

• Supports sequential, parallel, group-based digital signing workflows
• Supports Upload, View, Send, Sign, Audit, Download, Filtering and Delete document.
• Drag & drop signature placement with visual signature overlay
• Authorization with RBAC and authentication with JWT.
• Inlcudes Admin role.

A lightweight, self-hosted, deployless, anonymous public real-time chat app that becomes publicly accessible without deploying to a cloud server. Instead of deploying to a VPS or cloud platform, the application runs entirely on the host's local machine and uses Ngrok tunnel to expose it to the internet through a secure public URL.
Users can join existing chat rooms or create their own rooms with optional password protection. No registration, email, or user accounts are required, only a temporary username is needed to participate in conversations.

• Real-time messaging using WebSockets and utilizing Go goroutines concurrency and channels
• Create and join multiple chat rooms
• Anonymous usernames (no accounts, resgistration or authentication system)
• Self-hosted and publicly accessible through Ngrok tunneling
• SQLite for storing chat history
• Ngrok integration to expose the local server as a public URL without deploying to a VPS or cloud service

A lightweight HTTP server library for Go, built from scratch using TCP connections without relying on Go's http package, giving deep insight into http ptocol and network-level backend development.

• Implemented an HTTP server directly on top of Go's net TCP package.
• Built HTTP request parsing and response handling from the ground up.
• Designed a simple routing system supporting HTTP methods and custom handlers.
• Added static file serving for documents and media.
• Implemented configurable HTTP response headers and status codes.
• Added JSON and byte-based response support.
• Designed the library with a simple API for defining routes and starting the server.

A fast CLI tool see running processes and overview of a computer written in Go for Linux systems. Provides a clean overview of running processes and summary information of computer hardware and software.

• Monitors system processes and resource consumption based on RAM usage
• See information about computer hardware and software
• Display information about Operating System, RAM, CPU, DISK, Network, GPU and user
• Built specifically for Linux environments

A web-based OCR application for recognizing Javanese language from images and converting it into digital text. Built with Golang for the backend and Tesseract OCR as the recognition engine, featuring a clean and minimal interface with image preview, drag-and-drop upload, OCR processing, and downloadable text results.

• Built a web-based Javanese OCR application using Golang and Tesseract OCR.
• Implemented image upload, drag-and-drop, real-time image preview, OCR text extraction, and downloadable output
• Used tessdata_best model provided by the tesseract engine community.
• Containerized the app with Docker. Available at Docker Hub

A lightweight command-line utility for compressing and extracting files and folders using ZIP archives. Built with Go, Zipper provides a simple CLI interface for quickly zipping individual files, multiple files, or entire directories, as well as extracting ZIP archives to a specified destination.

• Zip individual files or multiple files or even entire directories
• Extract ZIP archives with custom destination paths
• Simple command-line interface
• Lightweight and fast

Built an end-to-end web application that detects multiple Indonesian freshwater fish species from a single image using my own YOLOv12 object detection model. Independently created the dataset, performed manual annotation and augmentation in Roboflow, trained the model in Google Colab, deployed it to a web application. The application identifies multiple fish and multiple species simultaneously, displays nutritional information for each detected species, and generates personalized cooking recipes using a Large Language Model (LLM).

• Multi-object and multi-species fish detection in a single image.
• Perform dataset collection, annotation and augmentation in Roboflow and trained the model in Google colab. Link to Roboflow ->
• Supports detection of Milkfish (Bandeng), Catfish (Lele), Tilapia (Nila), Pangasius (Patin), and Gourami (Gurame).
• Displays nutritional information for each detected fish species.
• AI-generated cooking recipes based on detected fish using an LLM.
• User-friendly web interface for image upload and real-time inference.