At Crowdgen Pro, I label and audit text, audio, and vision datasets used to train and fine-tune generative AI models. I validate work against evolving guidelines for consistency and semantic clarity.
I transcribe spoken audio into structured training text and label objects and bounding boxes for vision models. I also red-team generative agents with adversarial prompts to evaluate safety guardrails and unauthorized external tool actions.
At Jo's Java, I analyzed user engagement and digital interaction trends, contributing to campaign strategies that increased engagement by 35%. I also used ChatGPT and Claude to parse market insights and streamline structured reporting.
At Sierra Club (IL Chapter), I coordinated data-backed advocacy initiatives that increased community volunteer participation by 15%, created campaign communications, and organized fundraising events that generated $550.
