AI Data Annotation Specialist – Handshake AI (Project Hedgehog)
Worked on Project Hedgehog with Handshake AI, focusing on training and improving large language models through high-quality data annotation and evaluation. Tasks included Named Entity Recognition (NER), prompt and response writing (SFT), RLHF-based response ranking, and qualitative evaluation of AI-generated outputs for accuracy, coherence, and safety. Contributed to dataset refinement by identifying inconsistencies, bias, and factual errors. Maintained strict quality control standards, meeting weekly performance benchmarks and adhering to detailed annotation guidelines. The project involved reviewing and annotating thousands of text samples across multiple domains to enhance model reliability and contextual understanding