Capital One
Applied Researcher 4 (AI Foundations)
$219K–$272K
San Jose, CA +2 · Full-time · Posted Oct 14, 2026
- Location
- San Jose, CA
New York, NY
McLean, VA - Experience
- 2 years of Applied Research with a relevant M.S.; no years specified for the PhD route. Preferred NLP route includes a Master’s plus 5 years of industrial NLP research experience.
- Education
- PhD in a relevant field, current or in progress (degree obtained by scheduled start date), or M.S. in a relevant field.
- Visa sponsorship
- Capital One will consider sponsoring a new qualified applicant for employment authorization.
About the job
Conduct applied AI research and develop foundation models from design through deployment to power scalable customer-facing banking products.
What you'll do
- Build, train, evaluate, validate, and implement AI foundation models.
- Partner with cross-functional teams to deliver AI-powered products and translate research into business outcomes.
- Conduct applied research and collaborate on internal technical reports or conference submissions.
What they ask for
- PhD in a relevant field, current or in progress, or a relevant M.S. plus 2 years of Applied Research experience.
- Hands-on experience developing AI foundation models and solutions using open-source tools and cloud platforms.
- Experience building large deep learning models and delivering models at scale.
- Track record of machine-learning research contributions and ability to independently pursue a research agenda.
Nice to have
- Publications related to large language model pre-training or deep learning, including at ACL, NAACL, EMNLP, NeurIPS, ICML, or ICLR.
- Experience training a large language model from scratch or optimizing or deploying large models.
Pay
$218,700 - $249,600 USD base salary (McLean, VA); $238,600 - $272,300 USD base salary (New York, NY); $238,600 - $272,300 USD base salary (San Jose, CA)
Tools
Pytorch, AWS Ultraclusters, Huggingface, Lightning
AI focus
Applied research and development of AI foundation models, including large language models, training optimization, self-supervised learning, robustness, explainability, RLHF, and fine-tuning.
Model-written summary · the original posting is authoritative