Mert
Aytemiz
AI Software Engineer
I build large language model (LLM) systems: RAG, AI agents, and automated evaluation.
AI Software Engineer
I build large language model (LLM) systems: RAG, AI agents, and automated evaluation.
Profile
I make the architectural calls on large language model products, then measure whether they hold: how fast they answer, how often they are right.
Selected Work
01
A retrieval layer built on plain files, not a vector database. Answers as well as the RAG pipeline it replaced, on 95% fewer tokens.
02
Agents that choose their own tools through function calling, scored by an automated evaluation pipeline before release.
03
Token-by-token responses over SSE and WebSockets, tuned until the interface feels instant for businesses across the US and Türkiye.
04
The official Turkish localization of Prompt Engineering Guide, the 70,000-star open-source reference for working with LLMs.