Information Extraction Specialist

About the role

We are looking for an Information Extraction Specialist to design, build and maintain pipelines that transform unstructured data into structured, high‑quality datasets. The role is fully remote and involves collaborating with annotation teams, data scientists and domain experts across various industries.

Key responsibilities

• Design and implement extraction pipelines for documents such as legal contracts, medical records, financial reports, news articles and technical manuals.
• Create and manage high‑quality training datasets, define sampling strategies and oversee annotation teams.
• Evaluate model performance using precision, recall and F1, analyse errors and iterate on guidelines or model architecture.
• Assess, integrate and customize open‑source and commercial extraction tools and platforms.
• Develop scripts and automated workflows to streamline the extraction process.
• Adapt extraction systems to new domains, quickly acquiring domain knowledge and updating guidelines.

Required profile

• Minimum 5 years of experience in information extraction, NLP or computational linguistics.
• Strong written and verbal communication in English, with ability to document complex guidelines.
• Proven ability to design annotation schemas and handle edge cases.
• Experience evaluating extraction models and interpreting metrics such as precision, recall and F1.

Required skills

• Python programming
• spaCy
• NLTK
• Hugging Face Transformers
• Stanford CoreNLP