At micro1, I annotate and diarize audio and sound clips for AI training datasets, labeling speaker turns, sound events, and source characteristics according to project-specific guidelines.
I also review annotated clips, focusing on confirmed audible properties rather than visual assumptions or inferred emotion.
On the PP19 Project with Productive Playhouse / Gemini, I design and refine rubrics for evaluating AI-generated responses and content. I score submissions against rubric criteria and flag edge cases for calibration.
I support multilingual evaluation projects in English, French, and Kinyarwanda, including phone-based AI response assessment. My background also includes customer-facing and remote support roles where accurate documentation and structured workflows mattered.
