Mai Mayeg

I’m an AI/NLP Engineer with hands-on experience building end-to-end LLM applications, from data ingestion and annotation through model integration, prompt engineering, and full-stack deployment. I’m comfortable designing evaluation and benchmarking pipelines, especially for retrieval-augmented (RAG) components.

In my recent experience at Flink SE, I optimized end-to-end real-time IT asset tracking using Snipe-IT and Jira Assets, and I standardized internal troubleshooting knowledge into a reusable base that improved team response times. I also designed and prototyped an AI-powered shopping assistant that turned natural-language recipes into structured, automated shopping carts using LLMs and search APIs.

Earlier, at Comtravo, I annotated and curated Named Entity Recognition (NER) datasets in the travel domain and performed systematic validation/QA to ensure consistent, accurate model outputs. During my R&D internship at Daimler AG, I benchmarked and tested speech dialogue systems for in-car voice interaction (MBUX) and implemented Python scripts for information retrieval from textual data. I’m currently completing my M.Sc. in Cognitive Systems at the University of Potsdam, with a planned thesis on an LLM-based web application for rubric-aligned automated English essay feedback.