All Projects

An asynchronous AI-powered text processing REST API built with FastAPI and local LLM inference through Ollama. The service provides summarization, text classification, named entity extraction, rewriting, and translation through validated API endpoints. The project demonstrates a production-minded AI Engineering architecture by separating API routing, validation, AI services, and model providers. It includes asynchronous model communication, structured JSON generation, Pydantic validation, retry mechanisms with exponential backoff, structured error handling, and containerized deployment with Docker. The system uses Qwen 3 1.7B as the default local inference model, allowing text processing to run locally without relying on external AI APIs.

ATHLEEA APPAREL is a modern e-commerce platform built to deliver a seamless and visually engaging shopping experience for football enthusiasts. The website features a responsive interface, curated product collections, dynamic product discovery, detailed product pages, and a streamlined purchase flow that connects customers directly to WhatsApp for personalized checkout. Designed with a strong focus on performance, usability, and clean product presentation, Athleea demonstrates the ability to transform a real-world apparel business into a polished and production-ready digital storefront.

Bioskop CACB is a Python-based CLI cinema ticket management system designed to streamline movie booking, customer queuing, and transaction handling. It combines FIFO Queue and Linked List data structures with searching and sorting algorithms to provide structured queue management, interactive seat selection, persistent transaction records, and reliable order processing. The system also features automated PDF invoice generation and sales analytics, providing insights such as best-selling movies and busiest days. Robust input validation and error handling ensure a stable and interactive terminal experience throughout the ticketing workflow.