Marius Alin Staicu

Key Skills

Software

No software listed

Top Subject Matter

GPT-based AI evaluation

Software Asset Management & License Compliance

Business Process & Enterprise Compliance

Top Data Types

Text

Document

Top Task Types

No task types listed

Freelancer Overview

My professional background sits at the intersection of data quality, compliance analysis, and AI evaluation — a combination that translates naturally into AI training work. Over the past decade, I've built analytical frameworks at Oracle and E.ON focused on identifying errors, assessing risk, and maintaining accuracy in complex datasets. At E.ON, I tested and evaluated a GPT-based AI solution end-to-end, providing structured feedback on model performance and usability — work that required both critical thinking and a clear understanding of how LLMs behave in real-world contexts. More recently, I've been actively deepening my hands-on AI expertise through independent evaluation projects, working with models across writing quality, factual accuracy, instruction-following, and response calibration. I'm familiar with tools like ChatGPT, Microsoft Copilot, Google Gemini and Claude AI, and I approach AI outputs the same way I approached compliance analysis: systematically, with attention to edge cases and a focus on what's actually wrong rather than what looks good on the surface. That combination of structured analytical thinking and direct LLM experience is what I bring to AI trainer work.

Labeling Experience

Education

Work History