About

Machine Learning Engineer focused on LLMs, multi-agent systems, and production AI. Here is my background and how I work.

Background

I have 5+ years building and shipping ML systems, from edge devices to cloud APIs. I started in AI/ML at Data Pilot in Lahore (sound classification on Raspberry Pi), then moved to GagaMuller in Dublin building multimodal RAG and diffusion pipelines. At FZI Forschungszentrum Informatik in Karlsruhe I worked on computer vision and anonymization for autonomous driving. I am currently an AI Python Engineer at Schoene Neue Kinder in Munich, where I design multi-agent systems with LangGraph and conversational AI for internal tools and game developers.

I contribute to open-source (WhisperX, FastAPI tooling, RAG frameworks) and implement research such as HANRAG and KERAG. I am based in Koblenz, Germany, and open to relocation. Languages: English (C1), German (A2). When not coding, I enjoy tennis, gaming, and following AI research.

Education

Master of Science in Web & Data Science, University of Koblenz, Germany (Apr 2024 – ongoing). Relevant coursework: Big Data, Data Science, Foundation Models, Web Retrieval, Pattern Recognition, Network Theory.

Bachelor of Software Engineering, COMSATS University Islamabad, Pakistan (Sep 2017 – Aug 2021). Relevant coursework: Artificial Intelligence, Data Visualization, Software Requirement Engineering.

Skills

ML/AI: PyTorch, TensorFlow, Transformers, LangChain, LangGraph, Llama Index, RAG, Computer Vision, NLP.

Languages & frameworks: Python, TypeScript, Java, Node.js, R — FastAPI, Flask, NestJS, Streamlit.

Cloud & infra: AWS (Bedrock, SageMaker, S3), GCP (Vertex AI), Docker, MLOps, CI/CD.

Data: SQL, MongoDB, PostgreSQL, Redis.