Raifa Ananda

At Outlier’s Aether Project, I evaluate and rank LLM-generated responses for factuality, relevance, instruction following, logical consistency, and overall quality. I also verify claims and identify unsupported statements, inaccuracies, contradictions, and reasoning errors.

I contribute to RLHF workflows through comparative preference evaluation and response ranking. My work also includes image and audio annotation, visual content evaluation, image-text alignment, and ASR/TTS assessment for transcription accuracy, intelligibility, and speech naturalness.

As a native Indonesian speaker with a Philosophy background from the University of Indonesia, I bring careful attention to language, semantic nuance, context, logic, and critical inquiry across AI evaluation tasks.