I collaborate closely with development teams to design and execute creative, high-impact prompts that evaluate and stress-test generative AI models across text & image. My primary focus is on identifying model vulnerabilities, probing edge cases, and delivering detailed feedback to enhance model accuracy, safety, and reliability. Analyze and annotate model outputs to surface biases, hallucinations, or unsafe content. Contribute to the iterative refinement of generative models through continuous evaluation and structured feedback. Stay current with advancements in generative AI safety and prompt engineering best practices. Support the broader mission of bridging AI research with production-ready Meta products by ensuring alignment with safety policies. This role operates within Meta’s Generative AI organization, which focuses on translating cutting-edge AI research into practical, scalable solutions that power Meta’s products.

Perform quality evaluation of the automated results and labeling data. Analyzing data and providing feedback to engineers to help improve data quality. Identifying patterns from data and reporting to the team. Perform quality evaluation to measure the different aspects of the data and the automated exploration. Perform bootstrapping, login and configuration of some key apps to unblock the ability to explore these apps.

Rate and labeling data

Assure data quality and match the desired measurement

IT proficient in designed platform

English, Portuguese and Spanish languages

Knowledge in Maps, locale, and categorization

QA for awkward language in web ads or ads that poorly matched

In-depth knowledge of their culture to help select and classify website for the client

Improve the accuracy/relevancy of ads that are placed with search results and on websites

Quality assurance for Portuguese vocabulary, grammar for iphone’s Siri and texting functionality

Culturally awareness of current events, trends, slangs in Brazil

Proficient IT skills

Understanding how to track and report issues

Enhance and Improve the localization system, voice recognition technologies and maps in the Portuguese- Brazil language.