I am preparing a new round of model training and need a senior-level AI developer who can take full ownership of the data-preparation stage—especially data annotation. The core of the job is to design and implement a robust annotation workflow for a sizeable corpus of text data, then feed that clean, well-labeled material back into the training loop. You should already have hands-on experience setting up annotation guidelines, managing annotators or automation tools, and integrating the resulting labels into a machine-learning or deep-learning pipeline. Familiarity with popular NLP libraries (spaCy, Hugging Face, TensorFlow, PyTorch, etc.) will be essential, as the final objective is to boost downstream model performance by improving label quality and consistency. Deliverables • A documented text-annotation schema aligned with task objectives • A functioning annotation pipeline (can be human-in-the-loop, semi-automated, or fully automated—whichever you recommend and justify) • Clean, labeled text dataset ready for training • A brief report outlining best practices followed and next-step recommendations I value both speed and precision, so please mention any past projects where you have successfully combined those two qualities in a data-annotation context.