Staff AI Engineer - Remote & Foundation AI Systems

Capital One; Montpelier (VT); Remote; USD 245,000 - 279,000; Full time

Job summary

Capital One is hiring a Staff AI Engineer for a remote-eligible role. You will design and implement AI software components including foundation model training, LLM inference, agents and multi-model workflows, and governance, partnering with engineers, researchers and product managers.

You will optimize performance, latency, and cost using Open Source and SaaS tools such as AWS, HuggingFace, and PyTorch, while championing responsible AI across architecture, observability, and governance.

Qualifications

• Bachelor's in CS/AI/EE/CE or related field with 8+ years AI/ML experience, or Master's with 6+ years.
• At least 8 years of experience programming with Python, Go, Scala, CUDA, or Java

Responsibilities

• Deliver AI-powered products with cross-functional teams.
• Design, develop, test, deploy AI software components.
• Leverage Open Source and SaaS AI tech like AWS Ultraclusters, Huggingface, PyTorch.
• Invent foundation model optimization techniques for scale, cost, latency, throughput.
• Contribute to the technical vision and long-term roadmap of foundational AI systems.
• Set the technical direction for enterprise-wide AI architecture and standards.
• Own the design/integration of model routing, caching, and orchestration.
• Champion responsible AI principles across system design.
• Drive education, mentorship, and dissemination of best practices through councils.

Skills

Python

Go

Scala

CUDA

Java

ML systems

Education

Bachelor's degree in CS/AI/EE/CE or related fields

Tools

AWS

Google Cloud

Azure

Huggingface

PyTorch

VectorDBs

CUDA

Job description

Capital One is hiring a Staff AI Engineer for a remote-eligible role. You will design and implement AI software components including foundation model training, LLM inference, agents and multi-model workflows, and governance, partnering with engineers, researchers and product managers.

You will optimize performance, latency, and cost using Open Source and SaaS tools such as AWS, HuggingFace, and PyTorch, while championing responsible AI across architecture, observability, and governance.