Middle Applied Scientist (#5884)
- Рівень:
- middle
- Джерело:
- djinni.co
Що робити
- Partner with senior and principal data scientists to improve dataset discovery, qualification, and selection by evaluating and enhancing retrieval, classification, filtering, and ranking approaches
- Develop and apply evaluation and sampling methods to measure dataset relevance, coverage, diversity, composition, and representativeness.
- Evaluate automated dataset-curation workflows, using structured error analysis to assess interpretation, selection quality, and failure modes and recommend improvements.
- Design and analyse experiments using offline evaluation and product telemetry to recommend improvements to curation quality, workflow performance, and customer experience.
Що очікуємо
- 4+ years of commercial experience in data science, applied science, machine learning, or a related field.
- Strong grounding in statistics, experimental design, and quantitative problem-solving.
- Experience evaluating search, recommendation, ranking, and similar machine-learning systems.
- Experience taking experiments from problem framing and metric definition through analysis and recommendation.
- Proficiency in Python and SQL and experience working with large datasets.
Що пропонуємо
- Experience with embedding models, semantic search, information retrieval, and multimodal content.
- Experience evaluating LLM-based or agentic workflows.
- Understanding of annotation workflows, from human labelling to model-assisted and VLM-based annotation.
- Experience estimating full-population quality from representative samples.
Схожі вакансії
З блогу Trackr
Усі статті →
Знайдено через trackr.help/jobs · Канал: @trackrhelp · Бот для персональних сповіщень: @trackrhelpBot