About the role
As an AI Data Curator at TheWebVale, you will ensure our artificial intelligence models are trained and evaluated on impeccable data. You will organize, annotate, and sanitize specialized datasets for healthcare practices, educational institutes, and commercial retail brands.
You will collaborate with AI engineers and client domain experts during our regular two-to-three week sprint cadences. You will structure domain taxonomies, eliminate annotation bias, and maintain high standards of dataset documentation.
What you will do
- Collect, curate, and structure domain-specific text and document datasets for AI models.
- Write comprehensive annotation guidelines and rubrics for domain labeling tasks.
- Perform rigorous data hygiene, removing duplicates, artifacts, and personal identification data.
- Audit annotated datasets for inter-rater agreement and correct labeling discrepancies.
- Structure training, validation, and test splits to prevent data leakage across benchmarks.
- Document dataset provenance, licensing terms, and distribution characteristics.
What you will need
- Hold 2–4 years of experience in data curation, annotation management, or linguistic analysis.
- Demonstrate strong proficiency in Python, Pandas, regular expressions, and spreadsheet tools.
- Show practical experience working with JSON, JSONL, CSV, and Parquet data formats.
- Understand machine learning data requirements including balanced classes and edge cases.
- Possess exceptional attention to linguistic subtlety, tone, and factual precision.
- Communicate data guidelines and quality expectations clearly to annotators and engineers.
Nice to have
- Experience using annotation platforms like Label Studio, Prodigy, or Argilla.
- Familiarity with synthetic data generation techniques using large language models.
- Background in linguistics, technical writing, or library information science.