Arda Duman

Engineering Software-intensive Systems Major. Minors: Computer Architecture, Distributed Systems, Algorithms.

Passed with Merit (Gesamtnote 2.5) · Thesis grade: 1.3

GPA: 3.03/4.00 · High Honor List · 50% Merit Scholarship

• Computer Systems, Networks and Software
• Cryptography, Security, and Theory
• Data Analytics

• Delivered 60+ merged pull requests across feature development, UI redesigns, and bug fixes for a B2B mobility platform using Kotlin and Java, integrated via REST APIs with third-party services.
• Led the migration of 13 screens from XML to Jetpack Compose and built a reusable design system (typography, input, toggle, list components) adopted across the entire codebase.
• Refactored core modules to improve separation of concerns and long-term maintainability.
• Collaborated with product and design teams in Agile/Scrum sprints to translate enterprise requirements into scalable API designs and frontend solutions.

• Developed an internal energy trading platform using Angular, serving operations across multiple business units.
• Architected a reusable component library of 15+ modular UI components, reducing feature development time across teams.
• Improved UI load performance through CSS and data-binding optimization.

• Rebuilt the Setur Mobile Application from scratch using React Native and TypeScript, delivering a cross-platform foundation scaled to millions of users.
• Developed core user-facing screens including flight, hotel and tour search flows with filtering & sorting, reservation calendar, shopping cart, user registration, search history, and hotel listing cards.
• Led the migration from JavaScript to TypeScript, establishing type safety across a team of 5+ engineers.

• Built a Flutter mobile application for real-time ID card scanning, integrating ML-based OCR and classification services.
• Implemented RESTful API integration between the Flutter client and Python-based ML pipeline.

A toroidal grid puzzle game shipped to the App Store using Swift and SwiftUI. Features StoreKit 2 subscriptions and IAP, CloudKit leaderboards, Game Center achievements, and AdMob integration with ATT consent. Built on a zero-backend, maintenance-free architecture.

A SaaS platform for intercity bus companies in Turkey providing real-time competitor price tracking, route optimization, and dynamic pricing management. Includes Opti, an AI-based pricing model that predicts how ticket price adjustments affect occupancy rates. Accelerated by Kworks (Koç University).

Full-stack anomaly detection system built on Azure. Python ML backend with statistical time-series models, Firebase Firestore for data storage, Firebase Cloud Messaging for real-time push notifications, and a Flutter mobile client for alert delivery.

An end-to-end framework integrating satellite-based damage detection, GIS road network analysis, and optimization modeling to support post-disaster aid distribution planning. The system translates detected building damage into risk-aware road network costs, enabling accessibility-aware routing and improved POD placement.

• Developed a ResNet-based model for binary damage detection using satellite imagery.
• Converted building damage into road exposure penalties through GIS-based impact zones.
• Computed accessibility using least-cost paths on the weighted road network.
• Formulated a mixed-integer location–allocation model to optimize POD placement and demand assignments.