Create high-quality datasets for machine learning projects.
Dataset Curation Specialist
Data collection, cleaning, annotation, documentation
Best for
- ▸Building high-quality training datasets for LLM fine-tuning with proper annotation workflows
- ▸Implementing deduplication and quality filtering pipelines for large-scale web scraped data
- ▸Designing bias audit frameworks for computer vision datasets across demographic groups
- ▸Creating synthetic data generation strategies with human validation loops
What you'll get
- ▸Detailed data collection strategy with web scraping specifications, API rate limits, and provenance tracking requirements
- ▸Multi-stage cleaning pipeline with deduplication thresholds, quality filters, and PII removal protocols
- ▸Annotation taxonomy with inter-annotator agreement targets, quality control procedures, and labeling guidelines
Clear dataset requirements including target task, domain, quality metrics, annotation budget, and performance objectives.
Comprehensive dataset curation strategy with collection pipelines, cleaning workflows, annotation guidelines, quality metrics, and documentation templates.
What's inside
“You are a Dataset Curation Specialist. You identify and eliminate the failure modes that sink real-world ML projects: noisy labels, hidden distribution shifts, leakage, and fairness cliffs that only appear at scale. - Hunt for label noise as your primary threat; datasets fail not from missing exampl...”
Covers
Not designed for ↓
- ×Training machine learning models or implementing neural network architectures
- ×Real-time data processing or streaming analytics pipelines
- ×Database administration or production data warehouse management
- ×Business intelligence dashboards or executive reporting
SupaScore
88.73▼
Evidence Policy
Standard: no explicit evidence policy.
Research Foundation: 7 sources (4 academic, 2 official docs, 1 industry frameworks)
This skill was developed through independent research and synthesis. SupaSkills is not affiliated with or endorsed by any cited author or organisation.
Version History
v5.5 final distill
Pipeline v4: rebuilt with 3 helper skills
Initial release
Prerequisites
Use these skills first for best results.
Works well with
Need more depth?
Specialist skills that go deeper in areas this skill touches.
Common Workflows
LLM Fine-tuning Data Pipeline
End-to-end workflow from raw data curation through specialized fine-tuning dataset preparation to model training
© 2026 Kill The Dragon GmbH. This skill and its system prompt are protected by copyright. Unauthorised redistribution is prohibited. Terms of Service · Legal Notice