AI Engineer 4 (Vision model (VLM) customization experience)
Capital One5 days ago
Cambridge, MA, USA +3 moreStaff+
Base Salary
$215k - $246k/yr
Responsibilities
- Partner with engineers, research scientists, technical program managers, and product managers to deliver AI-powered products.
- Design, develop, test, deploy, and support AI components for foundation model training, LLM inference, agents, similarity search, guardrails, evaluation, experimentation, governance, and observability.
- Use AI technologies including AWS Ultraclusters, Hugging Face, VectorDBs, and PyTorch.
- Develop foundation-model optimization techniques targeting scalability, cost, latency, throughput, and hardware utilization.
- Own end-to-end architecture and contribute to the technical vision and roadmap for foundational AI systems.
- Define and maintain SLOs covering latency, uptime, and model performance drift.
- Optimize GPU/TPU utilization and model inference pipelines with infrastructure engineering.
- Lead technical reviews for AI system deployments, including security, data governance, and compliance.
- Mentor Principal and Senior Associates on scalable design, performance tuning, and research-to-production translation.
Requirements
- Bachelor's degree in Computer Science, AI, Electrical Engineering, Computer Engineering, or a related field plus at least four years of experience developing AI and ML algorithms or technologies, or a master's degree in one of those fields plus at least two years of relevant experience.
- At least four years of programming experience with Python, Go, Scala, CUDA, or Java.
- Preferred experience leading AI system development with cost, latency, throughput, and accuracy tradeoffs.
- Preferred six years of experience deploying scalable and responsible AI solutions on cloud platforms such as AWS, Google Cloud, Azure, or equivalent private cloud.
- Experience designing, developing, delivering, and supporting AI services.
- Experience with AI and ML technologies including LLM inference, similarity search, VectorDBs, guardrails, and memory using languages such as Python, C++, C#, Java, CUDA, or Golang.
- Experience optimizing training and inference software for hardware utilization, latency, throughput, and cost.
- Experience building agentic AI systems and workflows.
- Proficiency designing distributed systems for model training, evaluation, and online inference at petabyte scale.
- Experience defining AI model governance processes, including producibility, lineage tracking, and automated retraining schedules.
- Ability to influence architectural decisions across multiple AI product lines or platforms.
- Ability to understand current AI research and apply novel techniques in production.
Benefits
- Comprehensive health, financial, and other benefits supporting employee well-being, with eligibility varying by employment status and management level.
- This role is eligible for performance-based incentive compensation, including cash bonuses and/or long-term incentives.
- The posting states that applications will be accepted for a minimum of five business days.
- Capital One will consider sponsoring a new qualified applicant for employment authorization for this position.
About Capital One
Capital One is a U.S. consumer and commercial bank that offers credit cards, checking and savings accounts, auto financing, and lending to small and large businesses. It earns revenue from interest income and interchange/fees across its card and banking products, and develops cloud-based digital services. Founded in 1994, Capital One is publicly traded on the NYSE (COF) and is headquartered in McLean, Virginia.