About

I’m a Full Stack AI Engineer in Munich. I work across backend services, web applications and applied AI, and I like owning a change from the database through the API to the interface.

I started in June 2020 as a working student at the Turkish Maarif Foundation in Istanbul while studying Computer Science, building Kotlin-based Android features and internal employee portals. In February 2021 I became a Full Stack Developer there and worked on learning-management and education workflows, fee management for the foundation’s branches and the reporting that gave its headquarters a view of branch operations.

From September 2023 I worked remotely from Munich as a contract Full Stack Developer for AMO Bazar (AMO Travel). I built a Next.js e-commerce application integrated with the weclapp ERP and a WhatsApp customer-support assistant on Twilio and the OpenAI API, with a handoff to staff.

Since November 2024 I’ve been a Full Stack Developer at Aampere, working on the Dealer Platform for the European electric-vehicle market: NestJS services, PostgreSQL data models, queue-based jobs, Next.js features, LangChain and LangGraph agents that check listings, and an MCP server for test data and AI-assisted QA. The work page has the details.

I care about readable code and automated tests, because they make the next change easier, and about what happens after a release: being able to spot a problem and understand its cause.

Skills

  • Languages: TypeScript, JavaScript, Python, Kotlin, SQL
  • Backend and data: NestJS, Node.js, FastAPI, REST APIs, PostgreSQL
  • Frontend: Next.js, React
  • Applied AI: LLMs, RAG, AI agents, LangChain, LangGraph, MCP, OpenAI API
  • Cloud and tools: AWS (S3, CloudFront), Docker, Playwright, Vitest, Sentry, PostHog, Git, GitHub

Education

  • Bachelor of Computer Science, Software Engineering

    Kardan University · 2017–2021 · CGPA 3.8

  • German language course (B1)

    Kolping-Akademie, Munich · January – December 2024

Publication

A. Ghaffar, M. Akbari and M. Yousufzai, “Milad Scheduling Algorithm (MSA)”, Kardan Journal of Engineering and Technology, vol. 3, no. 1, pp. 110–116, December 2021. The paper proposes a CPU scheduling algorithm and compares its scheduling behavior with FCFS, SJF and Round Robin.

Languages

English, German (B1) and Persian (Farsi).