Igor Vodráška

I’m a film and media professional turned AI evaluation specialist, bringing 10+ years of audiovisual production, scene analysis, continuity, and editorial QA to multimodal AI projects. I combine visual interpretation with structured judgment so AI outputs are not just readable, but correct, useful, and guideline-aligned.

In my current work as an AI Language & Content Evaluator / LLM Response Evaluator, I validate AI-generated responses for linguistic quality, instruction-following, factual consistency, usefulness, clarity, and overall quality. I apply detailed guidelines and rubrics to identify quality issues, compare outputs, and write concise justifications for evaluation decisions.

My strongest differentiation is multimodal competence: I perform video/image annotation by describing subjects, settings, actions, emotions, temporal sequences, and interactions. From film practice, I also evaluate mood, lighting, spatial relationships, continuity, pacing, audience impact, and narrative logic—turning visual storytelling into reliable, structured descriptions.

Across long-term film roles (including First Assistant Director and Localization QA Director / Editor / Creative Producer), I’ve built the habits that make AI QA dependable: attention to detail, deadline reliability, and clear English communication. Whether I’m guideline-checking, hallucination-detecting, or supporting visual storytelling through editing and pacing, my focus stays the same—high-accuracy outputs with transparent reasoning.