Applied Researcher V

Capital One Financial

Confirmed live yesterday High trust

Quick summary

Work type: On-site
Location: New York, NYMcLean, VACambridge, MASan Jose, CA
Salary: $262,500–$299,600 / yr
Employment: Full-time
Posted: 5 days ago
Freshness: Confirmed live yesterday
Nearby: 99+ roles within 25 mi

Market check

Salary context

Above market

How this pay compares to similar roles

Similar $223k

This role $281k

$171k most similar roles pay here $313k

This role pays more than 95% of similar roles. Most pay $196,750–$248,550 — the shaded band above. At the midpoint, this role pays about $281k versus about $223k for comparable roles.

Based on 240 similar postings.

Employer

About Capital One Financial

Capital One Financial is a bank holding company specializing in credit cards, auto loans, banking, and savings products, known for its data-driven approach to consumer and commercial finance. Industry: Financial Services & Banking

Capital One Financial currently has 936 open roles on FindRole.

Listed pay typically runs $197,300–$225,100 across 933 roles with salary data.

Most-posted roles

View all roles at Capital One Financial

At a glance

TL;DR · Applied Researcher V

Role Posting Log in to save

Applied Researcher 5 (AI Foundations - LLM, Optimization and Finetuning) joins the AI Foundations team to develop trustworthy and reliable AI systems for banking services. This role involves collaborating with cross-functional teams of data scientists and engineers to build AI foundation models through every phase of development, including design, training, evaluation, validation, and implementation. The researcher will translate complex scientific insights into tangible business goals while managing research threads that bridge prototype development and production deployment. Key technical requirements include expertise in large deep learning models, training optimization, self-supervised learning, robustness, explainability, and RLHF. Candidates will utilize a technology stack including Pytorch, AWS Ultraclusters, Huggingface, and Lightning to analyze massive volumes of numeric and textual data. The work focuses on advancing the state of the art in AI to improve customer interactions with financial products.

Skills

LLM Pytorch AWS Huggingface Lightning Deep Learning NLP RLHF Self-Supervised Learning Quantization Model Sparsification Training Optimization Compiler Design Instruction-Tuning

What you'll do

Build AI foundation models through all phases of development including design, training, evaluation, and implementation.
Conduct high-impact applied research to integrate state-of-the-art AI developments into customer experience products.
Develop large deep learning models using PyTorch, Hugging Face, and other open-source tools on cloud platforms.
Optimize model training and inference through techniques like quantization, sparsification, and parallel training design.
Lead research threads that bridge the gap between prototype development and production deployment with engineering teams.
Translate complex technical findings into tangible business goals for cross-functional stakeholders.
Identify and solve large, undefined problems by researching and evaluating emerging technologies.
Own and execute a research agenda by selecting impactful problems and managing long-running projects independently.

What we're looking for

Must have a PhD in a relevant field with 2 years of experience or an MS with 4 years of experience in Applied Research.
Experience building large deep learning models for language, images, events, or graphs is required.
Expertise in training optimization, self-supervised learning, robustness, explainability, or RLHF is required.
Must have a track record of delivering models at scale for both training data and inference volumes.
Experience delivering libraries, platform-level code, or solution-level code to existing products is required.
Proven ability to own a research agenda and independently carry out long-running projects is required.
A track record of high-quality machine learning ideas or improvements, such as first-author publications, is required.
Specific expertise in LLM pre-training, optimization, or fine-tuning is preferred.