Principal Applied Scientist

About the role

As a Principal Applied Scientist at Oracle, you will drive the development of next‑generation Generative AI and Large Language Model (LLM) solutions for healthcare and enterprise customers. You will translate business needs into cutting‑edge AI projects and guide them from research proof‑of‑concept to production.

Key responsibilities

• Partner with product managers to turn business requirements into AI initiatives.
• Collaborate with technical leaders to deliver models and integrate services on schedule.
• Coordinate multinational teams to move projects from research POC to production deployment.
• Design and review AI solution architecture, covering data pipelines, model training, and evaluation.
• Lead and mentor junior and senior applied scientists.
• Write production‑grade code and champion best engineering practices.
• Participate in project planning, reviews, retrospectives, and risk mitigation.

Required profile

• Proven experience designing and scaling AI models for production environments.
• Deep knowledge of Machine Learning, Deep Learning architectures (e.g., Transformers) and advanced training methods.
• Hands‑on expertise with the latest LLM and Generative AI techniques such as parameter‑efficient fine‑tuning, instruction fine‑tuning, and prompt engineering.
• Experience with emerging LLM frameworks and plugins (LangChain, LlamaIndex, VectorStores, Retrievers, LLM Cache, LLMOps/MLflow).
• Track record of building data collection/annotation pipelines and systematic evaluation processes.
• Strong publication record in top‑tier AI conferences or journals.
• Leadership experience guiding senior and early‑career scientists.

Required skills

• Large Language Models (LLMs)
• Generative AI
• Machine Learning
• Deep Learning
• Transformers
• Post‑training model alignment
• Multi‑lingual model alignment
• Model evaluation strategies
• Parameter‑efficient fine‑tuning
• Instruction fine‑tuning
• Advanced prompt engineering
• LangChain
• LlamaIndex
• VectorStores
• Retrievers
• LLM Cache
• LLMOps (MLflow)