Staff, Data Science & Applied AI

Is a core technical contributor within the Enterprise Data & AI Solutions team, supporting Warner Bros. Discovery's global portfolio across Studios, Streaming, Linear Networks, Consumer Products, Games, and Direct-to-Consumer platforms. It is designed for a hands-on expert in applied data science who thrives at the intersection of statistical rigour, machine learning engineering, and business impact.

• Design, develop, and deploy statistical, predictive, and machine learning models across domains such as customer analytics, forecasting, personalisation, optimisation, and content performance.
• Apply advanced techniques including ensemble methods, gradient boosting, deep learning, NLP, time-series forecasting, and recommendation systems, and ensure model robustness through rigorous validation, monitoring, and performance tracking.
• Design and implement Generative AI solutions leveraging large language models (LLMs) for use cases such as knowledge retrieval, content intelligence, metadata enrichment, summarisation, and workflow automation.
• Develop and optimise prompt engineering strategies, evaluation frameworks, and guardrails to ensure high-quality, reliable outputs.
• Architect Retrieval-Augmented Generation (RAG) pipelines integrating structured and unstructured enterprise data sources.
• Fine-tune or adapt foundation models where appropriate using parameter-efficient techniques (e.g., LoRA, adapters) aligned with business needs.
• Implement evaluation pipelines to measure hallucination rates, bias, latency, cost efficiency, and model quality in production environments.
• Collaborate with Responsible AI and Governance teams to ensure compliance with enterprise AI policies, data privacy standards, and ethical AI practices.
• Collaborate with Data Engineering and DevOps teams to productionise ML and GenAI solutions in scalable cloud environments, design CI/CD pipelines for model lifecycle management, including experimentation tracking, versioning, and automated retraining
• And implement monitoring frameworks for model drift, prompt drift, performance degradation, and data integrity.
• Develop reusable ML and GenAI frameworks, accelerators, and internal utilities that improve productivity across teams, and advance automation initiatives to reduce manual workflows and enhance analytical velocity.
• Stay current with cutting-edge advancements in foundation models, multimodal AI, and agentic architectures to continuously elevate enterprise AI capabilities.

• Bachelor's degree, MS, or greater in Computer/Data Science, Engineering, Mathematics, Statistics, or related quantitative discipline.
• 8+ years relevant experience in data science, with 2+ years experience in GenAI.
• Demonstrated track record of delivering production-grade AI/ML solutions with measurable business impact.
• Hands-on experience designing and deploying Generative AI solutions using large language models (e.g., GPT-class models, open-source foundation models, or enterprise LLM platforms).
• Strong proficiency in prompt engineering, structured output design, few-shot learning strategies, and systematic prompt optimisation.
• Experience building Retrieval-Augmented Generation (RAG) pipelines integrating vector databases and enterprise data sources.
• Familiarity with embedding models, semantic search, and vector stores (e.g., Pinecone, Weaviate, OpenSearch, FAISS, or equivalent).
• Experience fine-tuning or adapting foundation models using parameter-efficient approaches (e.g., LoRA, adapters, instruction tuning).
• Understanding of LLM evaluation methodologies, including hallucination detection, bias assessment, response quality scoring, and cost-performance trade-offs.
• Experience working with enterprise AI platforms (e.g., AWS Bedrock, Azure OpenAI, Databricks Model Serving, Snowflake Cortex, or equivalent).
• Understanding of Responsible AI principles, data privacy considerations, and model governance requirements in regulated environments.

• Exposure to multimodal AI (text, image, audio, video) and agent-based workflows.