At Invisible Technologies, I review video annotations and assess video-to-prompt and image-to-prompt accuracy against gold-standard benchmarks. I also label products and tag locations in social media imagery for computer vision model training.
I evaluate text outputs for factual accuracy, coherence, and policy alignment, and document edge cases and hallucinations for QA and engineering teams. I work independently under strict confidentiality, data security, and compliance requirements.
At RWS Group, I annotate and review training datasets and evaluate AI-generated outputs across text, image, and multimodal tasks. I apply structured feedback methods to support reinforcement learning workflows.
