GitHub Resume

Welcome to @ahikmah/lab

A virtual sandbox for software development experiments, machine learning projects, system designs, and everything in between.

Research Log

Cloud & Architecture

Spring AI & MCP Architecture: Technical Feasibility Study

A proof-of-concept implementing Model Context Protocol (MCP) in Java Spring ecosystem, enabling a modular AI architecture where LLM intelligence is decoupled from backend execution tools.

Spring AIMCPJava 21+2 more
Software Engineering

Google Play Store Review Scraper

A production-ready web scraper for extracting app reviews using Selenium, with modern Python best practices.

PythonSeleniumWeb Scraping+1 more
Machine Learning

End-to-End Face Detection & Recognition System

A comprehensive computer vision system that first detects faces in images and then identifies them using deep learning embeddings.

Computer VisionDeep LearningPython+1 more
Machine Learning

Heart Disease Classification

Predicting heart disease using clinical parameters with 88%+ accuracy using Logistic Regression and Random Forest.

Scikit-LearnPandasClassification+1 more
Cloud & Architecture

Multiclass Text Classification on AWS SageMaker

End-to-end ML pipeline using DistilBERT for news article classification, deployed on AWS SageMaker with Lambda serverless inference.

AWS SageMakerHugging FaceDistilBERT+2 more
Machine Learning

Choice-Based Conjoint Analysis

Market research technique using discrete choice modeling to understand consumer preferences for Netflix subscription features.

StatisticsMarket ResearchPython+1 more
Machine Learning

LSTM-Based Customer Feedback Classification in Indonesian

This study implements LSTM neural networks with FastText word embeddings for multi-class classification of Indonesian customer reviews into categories: Account, Application, Customer Service, and Transaction.

LSTMDeep LearningNLP+3 more
Software Engineering

PixelCraft: AI-Powered Image SaaS

A full-stack AI SaaS platform for image manipulation (restore, recolor, remove objects) built with Next.js 14, Cloudinary AI, and Stripe.

Next.js 14TypeScriptSaaS+2 more
Machine Learning

Sonar-Based Rock vs Mine Detection

This study applies supervised machine learning to sonar signal data to distinguish between underwater rocks and mines. Logistic regression was selected for its robustness on small, high-dimensional datasets.

Logistic RegressionSonar Signal ProcessingBinary Classification+1 more
Machine Learning

NeuroDesk: AI Helpdesk Automation

An intelligent helpdesk ticketing system featuring automated issue classification using LSTM and FastText models to streamline support workflows.

NLPPyTorchDjango+2 more
Software Engineering

Bemco Melawan Corona

An educational 2D platformer game designed to teach children about COVID-19 prevention through interactive gameplay and quizzes.

Game DevelopmentConstruct 2Education+1 more