AI Safety Evaluator | Senior Trust & Safety Reviewer

I am a Senior Trust & Safety and AI Safety Evaluator with hands-on experience reviewing, annotating, and quality-checking LLM-generated content in Spanish and English. I have proven expertise in AI safety evaluation, red teaming, adversarial testing, step-by-step reasoning assessment, and policy alignment. I possess strong judgment, consistency, and analytical writing skills, with a demonstrated ability to document clear rationales and decision paths. I am comfortable working with sensitive, explicit, or potentially harmful content in high-stakes AI data environments. I have prior experience collaborating with AI Data Services companies and foundation model evaluation projects.

In my current role as a Language Data & Quality Reviewer at Volga Partners, I review and quality-check LLM-generated content in Spanish to ensure accuracy, logical consistency, and compliance with safety policies. I identify conceptual errors, methodological issues, and hallucinations in AI responses, and evaluate and label sensitive content including violence, hate/harassment, sexual content, misinformation, and illegal activities. I produce clear, reproducible rationales supporting safety decisions and ratings, and compare and rank multiple AI responses based on correctness, clarity, and risk level.

Previously, I worked as an AI Trainer and AI Content Evaluator at Outlier, where I trained and evaluated AI models through prompt and output review in Spanish. I assessed step-by-step reasoning quality and explanatory clarity of model responses, detected safety risks, bias, and policy misalignment in generated content, and rewrote and optimized outputs to improve accuracy, safety, and alignment.

I also have experience as a Spanish Instructor in a bilingual environment at True Fluency Kids, delivering engaging Spanish lessons for children using a communicative approach. I maintained ongoing communication in English with parents and academic staff and adapted linguistic content to diverse cultural contexts.

In addition to my AI and educational roles, I have relevant experience in customer service and sales involving sensitive case handling and client communication, as well as social event photography requiring discretion, ethical judgment, and confidentiality.

I am proficient in Spanish (native) and English (advanced), and skilled in using tools such as ChatGPT for advanced output evaluation, Gemini for comparative response analysis, Perplexity for fact-checking and verification, and various AI data annotation and labeling tools. I am available for remote contract roles, working hourly with high confidentiality and comfort handling sensitive content.