Applied Researcher I

About the role

The AI Foundations team at Capital One is central to bringing our vision for AI at Capital One to life. Our work spans the entire research lifecycle, from collaborating with academia to building production systems. We work closely with product, technology, and business leaders to apply the latest advancements in AI to our business.

Responsibilities

• Partner with a cross-functional team of data scientists, software engineers, machine learning engineers, and product managers to deliver AI-powered products that enhance customer interactions.
• Leverage a diverse stack of technologies, including PyTorch, AWS UltraClusters, Huggingface, Lightning, and VectorDBs, to uncover insights from vast amounts of numeric and textual data.
• Build AI foundation models throughout the development lifecycle, from design to training, evaluation, validation, and deployment.
• Engage in high-impact applied research to advance the next generation of customer experiences.
• Communicate complex AI concepts to stakeholders to align with business goals.

Requirements

• Currently has, or is in the process of obtaining, a PhD in Electrical Engineering, Computer Engineering, Computer Science, AI, Mathematics, or related fields, or a Master's degree in Electrical Engineering, Computer Engineering, Computer Science, AI, Mathematics, or related fields with 2 years of experience in Applied Research.
• Knowledge of emerging technologies and a strong track record of evaluating and applying them.
• Experience with open-source languages and tools, particularly in deep learning and AI.
• Experience building large-scale deep learning models and expertise in areas like training optimization, self-supervised learning, robustness, explainability, RLHF.
• A proven ability to deliver models at scale, both in terms of training data and inference volumes.
• Experience in delivering libraries, platform-level code, or solution-level code to existing products.
• A history of publishing in top-tier conferences and journals in machine learning and AI.
• Experience with large-scale deep learning-based recommendation systems and production real-time/streaming environments.
• Experience with common open-source frameworks such as PyTorch-Geometric, DGL, and contributions to open-source libraries for data quality, dataset curation, or labeling.

Qualifications

• Basic Qualifications: PhD in Computer Science, Machine Learning, Computer Engineering, Applied Mathematics, Electrical Engineering or related fields, or equivalent combination of education and experience.
• Preferred Qualifications: PhD in Computer Science, Machine Learning, Computer Engineering, Applied Mathematics, Electrical Engineering or related fields, with a focus on Natural Language Processing (NLP), multiple publications on pre-training of large language models, training large language models from scratch, or scaling graph models to greater than 50 million nodes.

Skills

• Innovative and creative problem-solving skills.
• Strong communication and interpersonal skills to translate technical concepts into actionable business goals.
• Technical proficiency in deep learning methodologies and tools.
• Experience with large-scale distributed systems and cloud computing platforms.

Benefits

Capital One offers a comprehensive benefits package, including health, financial, and wellness programs designed to support your total well-being. Learn more at the Capital One Careers website.

Pay

Salaries for this role range from $218,700 to $272,300 annually, depending on location. For more details, visit the Capital One Careers website.

Schedule

This role is full-time and permanent.

Behavioral Models

Experience with geometric deep learning, training models on graph and sequential data structures, and working on large-scale deep learning-based recommender systems.

Optimization

Experience with training optimization techniques, including model sparsification, quantization, training parallelism/partitioning, design, gradient checkpointing, and model compression.

Finetuning

Experience with supervised finetuning, instruction-tuning, dialogue-finetuning, parameter tuning, and transfer learning, model adaptation, and model guidance.

Data Preparation

Experience with tokenization, data quality, dataset curation, and labeling, and contributions to major open-source corpora and open-source libraries for data quality, dataset curation, or labeling.