BACKEND SOFTWARE ENGINEER
Istanbul → Munich → Istanbul
I build the systems behind the screens — APIs, data pipelines and services that scale.
SCROLL THE JOURNEY ↓THE JOURNEY
Computer Engineering at Boğaziçi, graduated with honors. Shipped my first real software on Arkenstone, an Android social-trading app — and closed the degree with a project I'm still fond of: an attention-augmented CNN for facial-expression recognition, reaching 72% accuracy.
Microservices and REST APIs in Java, Kotlin and Spring Boot at Doğuş Technology; then a year at INTERTECH building open-banking APIs on C#, .NET Core and Azure — secure coding under real financial regulations, CI/CD pipelines, and clients who notice every minute of downtime. Earned the Deep Learning Specialization along the way.
Moved to Munich for a master's in Data Engineering & Analytics at TUM. Built Selphspace, a serverless psychotherapy platform where patients log their day and therapists track progress — my part: extracting handwritten text with Textract, S3, DynamoDB and Amplify. Then led the Meta team on CGI DigitalRadar: 20 developers, four teams, and me learning to distribute tasks, resolve cross-team dependencies and present to the customer.
Two years with Heidelberg-based Everest Systems: ERP platforms in React and TypeScript, AWS solutions tuned for performance and scalability — enterprise software discipline with cross-functional teams.
A year deep in my master's thesis: integrating AI into legacy software systems — where data compatibility, performance and cybersecurity meet decades-old code. Graduated February 2025.
Back home, at Etstur: building and maintaining a high-volume invoice generation system for the accounting platform — Java, Spring Boot, PostgreSQL, Oracle and PL/SQL, where financial data leaves no room for 'almost correct'.
SELECTED WORK
A digital psychotherapy platform connecting patients and therapists, built serverless on AWS with a React frontend. I designed the handwritten-text extraction pipeline: Textract, S3, DynamoDB, Amplify.
VISIT ↗An event-management platform proposed by CGI, built by 20 students in four teams. As Meta team lead I owned the core structural components, cross-team coordination and customer presentations.
Master's thesis at TUM: how to integrate AI into legacy software systems without breaking them — data compatibility, performance and cybersecurity as first-class constraints.
Graduation research at Boğaziçi: a CNN architecture augmented with an attention mechanism for facial-expression recognition — 72% accuracy on FER2013/AffectNet, built with Python, TensorFlow and Keras.
A social-trading platform for Android, built as a two-course software-engineering group project at Boğaziçi. Java and Kotlin: UI components, maps and API integrations.
TOOLBOX
Java · Kotlin · TypeScript · JavaScript · C#
Spring Boot · .NET Core · React
PostgreSQL · Oracle · PL/SQL · MongoDB
AWS · Azure · Docker · Jenkins · Git · Gradle · Maven
JUnit · Jest · Mockito
Python · TensorFlow · Keras
Speaks English, Turkish and German.
Certified: Deep Learning Specialization — deeplearning.ai (2020).
Off the keyboard: football, swimming, basketball.
CONTACT
Open to interesting backend problems. The fastest way to reach me: