LLM Trainer
I collaborated with Turing’s LLM partner to create high-quality datasets for model fine-tuning and benchmarking. I was responsible for annotating robotic task videos by translating actions into detailed, timestamped descriptions to support LLM training. I also evaluated task success, identified errors, and ensured annotations complied with strict quality and consistency standards. • Reviewed and refined video annotations for clarity and adherence to guidelines. • Produced gold-standard datasets to enhance robotics and AI model performance. • Worked with internal or proprietary annotation tools to handle video data. • Supported the development of robust LLMs for robotic task understanding.