Narrative intelligence: advance identification and clustering of related content using embeddings, graph-based clustering and human-labelled evaluations.
AI-powered scoring signals: experiment with and integrate new signals including LLM-based relevance classification and source credibility.
Trust and integrity: develop approaches to identify coordinated activity, impersonation, scams.
Data quality and reliability: build automated validation, monitoring and safeguards.
Own agreed deliverables end to end, design experiments, contribute production code.
Requirements
5+ years of applied data science experience, deploying and monitoring models or scoring systems in production.
Strong Python and SQL expertise with large-scale, complex datasets.
Practical experience with LLMs: classification, prompt engineering, structured outputs, model evaluation.
Experience with clustering and graph-based techniques, embeddings, similarity measures.