AI Engineer 4 (AI Safety and Multimodal)
Core
Design, develop, test, deploy, and support AI software components including foundation model training, large language model inference, agents, multi-agent workflows, similarity search, guardrails, and model evaluation to build responsible and scalable AI systems for banking.
Role type
Senior IC AI Engineer (AI Safety and Multimodal)
Builds
Proprietary AI platforms and services empowering Capital One products with foundation models, LLMs, and agentic workflows.
Domain
Financial Services / Applied AI / Multimodal Systems
Deliverable
production ML models
Required skills
Foundation model training, LLM inference, agentic workflows, similarity search, safety guardrails, model evaluation, observability, system architecture, SLO definition, GPU/TPU optimization, AI red teaming, multimodal safety (VLMs, MLLMs, MM-RAG), distributed systems design
Preferred skills
Leading development with cost/latency tradeoffs, deploying scalable AI on cloud platforms, optimizing training/inference software, AI safety alignment, vulnerability assessment of multi-modal applications, building agentic AI systems, defining AI model governance processes
Technologies
AWS Ultraclusters, Huggingface, VectorDBs, PyTorch, CUDA, Python, Go, Scala, Java, C++, C#
Responsibilities
Partner with cross-functional teams to deliver AI-powered products; Design and deploy AI components including foundation models and agents; Invent optimization techniques for performance and scalability; Own end-to-end architecture for complex AI systems; Define and maintain SLOs for AI reliability; Collaborate on GPU/TPU optimization; Lead technical reviews for security and compliance; Mentor Principal and Senior Associates
Seniority
Senior, hands-on IC
