Company Description:
DeepLLMData specializes in providing high-quality human-generated data to power frontier AI and large language models. The organization collaborates with experts in coding, STEM, mathematics, and other domain-specific areas to support supervised fine-tuning and reinforcement learning from human feedback (RLHF). DeepLLMData also works with image and video annotators and manages comprehensive image and video data provisioning pipelines. Team members contribute to cutting-edge AI research and product development by creating and curating precise, well-structured datasets.
Role Description:
This is a remote, contract role for Video creators from US Region only for creating short language-focused video content for LLM model training. We are looking for short videos of 10 minutes duration with different individuals with following languages :Spanish EnglishPortugueseHindiKoreanChineseArabicJapaneseFrenchIf you have a Digital Camera for recording or a Tripod with Camera to create a still video, please apply.
Technical & Content Standards:
Format: 1080p, H.264 / MP4, Landscape orientation, 25-60 FPS.Environment: Plain white background; stable, controlled studio environment.
Speaker: One audible speaker only. No hand gestures or body movement while speaking.
Speech: Natural, unscripted (no reading allowed). Topic must be safe (e.g., hobbies, work, travel); religion and politics are strictly prohibited.
Quality: No unnatural pauses (>5s). Geotagged/timestamped proof of recording required for every submission.
