Nathnael Keleme

MenuFlow is a full stack digital restaurant platform built to modernize how restaurants manage menus, orders, staff operations, and customer experiences through QR based ordering and real time restaurant management. The platform was designed to solve common operational problems restaurants face daily, including slow service, order miscommunication, expensive menu reprinting, inefficient waiter coordination, and the lack of actionable business insights.

Customers can simply scan a QR code placed on a table to instantly access a digital menu with rich food images, pricing, and descriptions without downloading an app or creating an account. They can browse categories, add items to cart, and place orders directly from their phones, creating a faster and more seamless dining experience while reducing wait times and manual errors.

On the operational side, MenuFlow provides dedicated dashboards for both restaurant administrators and waiters. Waiters receive live order notifications in real time through WebSocket powered updates, manage customer orders, track shift activity, and create orders directly on behalf of customers when needed. Restaurant administrators can manage menu items, categories, tables, QR code generation, staff accounts, restaurant settings, and complete order workflows from a centralized dashboard.

The system also includes analytics and reporting tools that help restaurant owners monitor revenue trends, peak business hours, best selling items, and overall restaurant performance, enabling smarter business decisions driven by real operational data.

Built with a production ready monorepo architecture, the platform uses Next.js, NestJS, PostgreSQL, Socket.IO, Drizzle ORM, Redis, and Supabase Storage. It integrates secure JWT authentication, role based access control, rate limiting, responsive UI design, real time communication, and scalable backend workflows optimized for modern restaurant environments.

MenuFlow demonstrates scalable full stack engineering, real time application architecture, digital transformation for hospitality businesses, and user centered system design focused on improving restaurant efficiency and customer satisfaction.

AASTU Campus Event Management System (CEMS) is a centralized, AI powered platform built to modernize how campus events are organized, discovered, and managed at Addis Ababa Science and Technology University. The project was created to solve a common challenge across university campuses where event communication relied heavily on scattered Telegram posts, physical notice boards, and word of mouth, causing students to miss opportunities while organizers struggled to measure engagement and manage attendance effectively.

The platform connects students, organizers, and administrators through a unified digital ecosystem that streamlines the entire event lifecycle from event creation and discovery to attendance tracking and post event analytics. Organizers can publish events, manage registrations, upload promotional content, monitor participation metrics, and handle ticketing workflows, while students can explore upcoming events, receive personalized recommendations, RSVP, and check in securely using QR based attendance verification.

One of the core highlights of the system is its AI powered recommendation engine, which delivers personalized event suggestions using hybrid recommendation techniques based on user interests, attendance history, interactions, and engagement behavior. The goal was not only to improve event visibility but also to increase student participation and strengthen campus community engagement.

The system also includes real time notifications, digital ticket generation, QR code scanning, event approval workflows, schedule conflict validation, analytics dashboards, speaker and agenda management, role based access control, and integrated support ticket management. Additionally, a Telegram bot workflow was designed to simplify graduation event ticket distribution for parents and guests who may not have access to traditional email based systems.

Built with a production focused architecture, the platform uses NestJS with a Modular Monolith architecture, PostgreSQL, Redis, BullMQ, WebSockets, Next.js, Flutter, Docker, and AWS deployment infrastructure. The project demonstrates scalable full stack engineering, AI driven personalization systems, real time application design, and enterprise level software architecture focused on solving real problems within university communities.

Portfolio CMS Platform is a full-stack, production ready portfolio management system designed to help developers, designers, and professionals manage and showcase their work dynamically through a modern public portfolio website and a powerful real-time admin dashboard. The project was inspired by a common challenge faced by many software developers, continuously updating portfolios with new projects, certifications, hackathons, awards, trainings, testimonials, and professional achievements as their careers evolve. Instead of manually editing static portfolio code for every update, this platform provides a fully dynamic content management experience built specifically for personal branding and professional growth.

The system combines a visually rich, SEO optimized portfolio frontend with a secure CMS dashboard where all portfolio content can be managed in real time. Users can create, edit, reorder, publish, schedule, and organize portfolio sections including About Me, Skills, Experience, Education & Training, Projects, Awards & Recognition, Testimonials, Social Links, Resume management, and contact messages, all without modifying source code manually.

The platform includes advanced features such as drag and drop content ordering, rich text editing with Tiptap, image upload and media management, content versioning, backup and restore functionality, real time contact notifications via Socket.IO, audit logging, analytics dashboards, and global search across all portfolio content. The public facing portfolio delivers an interactive user experience with animated project showcases, image carousels, testimonial sliders, floating navigation, responsive layouts, and optimized performance for modern devices and search engines.

Built using a modern monorepo architecture with Next.js 14, NestJS, PostgreSQL, Redis, Socket.IO, and Drizzle ORM, the project emphasizes scalability, maintainability, and production grade engineering practices. It demonstrates enterprise level full-stack development principles including modular architecture, reusable shared packages, real time systems, secure authentication workflows, scalable CMS design, and developer-focused tooling, all tailored toward solving the real world problem of maintaining a constantly evolving professional portfolio efficiently and dynamically.

Food Delivery Aggregator is a production grade microservices based food delivery platform designed to support real-time restaurant ordering, payment processing, delivery management, and live customer notifications at scale. The system demonstrates modern distributed system architecture using independently deployable services, asynchronous messaging, real-time communication, and secure authentication workflows.

The platform consists of multiple specialized services including authentication, order management, payment processing, notification handling, API gateway routing, and frontend applications for customers, restaurants, drivers, and administrators. Each service owns its own database following the database-per-service architecture pattern, ensuring loose coupling, scalability, and service independence.

The architecture combines synchronous HTTP communication through Nginx reverse proxy routing with asynchronous event driven communication powered by RabbitMQ. Real time notifications and live order updates are delivered using Socket.IO WebSockets, enabling instant communication between customers, restaurants, and delivery personnel. The notification system consumes events from RabbitMQ queues and pushes live updates directly to connected users.

Security is implemented using JWT authentication with role based authorization, secure Argon2 password hashing, refresh-token workflows, email verification, and protected token storage strategies. The platform also incorporates resilience patterns including circuit breakers, retry mechanisms, graceful degradation, and connection recovery strategies to improve reliability and fault tolerance in distributed environments.

The project demonstrates enterprise level backend engineering practices, scalable infrastructure design, event driven architecture, and production-ready microservices communication patterns suitable for large scale real world delivery systems.

It's a comprehensive Gym and fitness management platform built to streamline daily gym operations, member engagement, attendance tracking, and business analytics within a single modern application. Designed with scalability and maintainability in mind, the system provides a complete solution for managing members, trainers, fitness classes, equipment, payments, and gym performance analytics through an intuitive and responsive interface.

The platform features advanced QR-based member check in with real time payment verification, ensuring only active members gain access to facilities and scheduled classes. Administrators and staff can efficiently manage memberships, monitor attendance, organize trainer schedules, track equipment maintenance, handle subscriptions and billing, and generate detailed reports for revenue and operational insights.

The project follows a feature-first architecture with clear separation between Presentation, Domain, and Data layers, making the system highly maintainable and ready for future backend integration or large scale deployment.