At Mindrift, I annotate, categorize, and label text and digital data according to project guidelines. I also verify records and document data-quality edge cases with reproducible steps.
At Toloka AI, I evaluated and ranked LLM outputs for accuracy, safety, helpfulness, and nuance. I also authored prompt-response pairs and categorized text datasets for generative AI.
At OneForma / Centific, I reviewed digital content against quality and safety guidelines, flagged edge cases, and transcribed audio and video datasets.
My earlier work included scheduling, email and calendar management, research, and data entry as a Virtual Assistant, as well as customer support at Teleperformance, where I maintained a 95%+ satisfaction rating.

