At Project Folio, I progressed from AI Data Annotator to Reviewer, checking AI training annotations for accuracy, consistency, and compliance with project guidelines. I verified bounding boxes, classifications, transcriptions, formatting, and reading order, and shared corrective feedback on recurring issues.
In AI safety and evaluation projects, I tested model responses with structured prompts and assessed them against safety, policy, relevance, and quality requirements. I documented problematic behavior and provided structured feedback to support improvements.
I’ve also worked on Odia language data, including annotation, transcription validation, and linguistic quality review. Through Mercor’s Introduction to Agents training, I applied structured evaluation principles to AI-generated outputs and agent behavior.
