Source description
About the role
Responsible for analysing and understanding of the massive structured and unstructured data by NLP tasks, such as data processing, NER, relationship extraction, NOR etc.
Own and improve our information-extraction pipeline, from model to production
Solve document-level extraction at scale — design how we handle long documents where entities, properties, and their relationships span across sections
Independently drive algorithm optimization, develop and fine-tune models to meet business requirements.
Build hybrid systems combining NER, rule-based methods, and LLMs, applying each where it genuinely wins
Define evaluation methodology / annotation strategy / quality metrics with domain experts
Process and extract from large-scale corpora reliably and efficiently
More at Patsnap
