Capital One Principal Associate Data Science - Applied Generative AI Calls Documents 

United States, Virginia, Arlington 

392182572

Center 1 (19052), United States of America, McLean, Virginia Principal Associate , Data Science - Applied Generative AI for Calls and Documents

In this role, you will:

• Harness the power of transformer model architectures to automatically identify emerging customer pain points in millions of call transcripts.

• Fine-tune large language models (LLMs) and large multi-modal models (LMMs) for extractive and abstractive tasks to search for complex evidence statements in unstructured, multi-page, documents images.

• Manage large scale data annotation projects by guiding frontline agents to curate high quality datasets, delivering model improvements by proposing, managing, and monitoring improvements to data collection processes.

• Work on a team of data scientists to build practical machine learning solutions through all phases of development, including designing, training, evaluating, and monitoring models.

• Communicate frequently with business stakeholders, including everything from brainstorming verbiage to include in a prompt engineering experiment to ascertaining which model evaluation metric best aligns data science outputs with business objectives.

• Collaborate with machine learning engineers to develop, deploy, troubleshoot, optimize, and maintain model pipelines with activities spanning from building reusable Kubeflow components for LLM fine-tuning to conversing about the cost impacts of model architecture choices.

• Leverage a broad stack of technologies including Pytorch, Hugging Face, LangChain, LLaMA-Factory, GitHub, AWS and more, to automate workflows using huge volumes of text audio, and vision data
  

The Ideal Candidate is:

• Creative. You thrive on bringing definition to big, undefined problems. You love asking questions and pushing hard to find answers. You’re not afraid to share a new idea.

• Innovative. You continually research and evaluate emerging technologies. You stay current on published state-of-the-art methods, technologies, and applications and seek out opportunities to apply them.

• Technical. You’re comfortable with open-source languages and are passionate about developing further. You have hands-on experience developing data science solutions using open-source tools and cloud computing platforms.

• Statistically-minded. You’ve built models, validated them, and backtested them. You know how to interpret a confusion matrix or a ROC curve.

• Passionate about the applied use of data science - when you see a new generative model take the top spot on a HuggingFace model leaderboard, you are just as excited about how it can improve a business process as you are about the underlying technical innovations.

• You have an ownership mindset for all upstream and downstream impacts to model pipelines. You like to question what imperfections exist in a model benchmark, taking self-initiative to fix data quality issues in evaluation data.

Basic Qualifications:

• Currently has, or is in the process of obtaining a Bachelor’s Degree plus 5 years of experience in data analytics, or currently has, or is in the process of obtaining a Master’s Degree plus 3 years of experience in data analytics, or currently has, or is in the process of obtaining PhD, with an expectation that required degree will be obtained on or before the scheduled start date

• At least 1 year of experience in open source programming languages for large scale data analysis

• At least 1 year of experience with machine learning

• At least 1 year of experience with relational databases

Preferred Qualifications:

• At least 2 years of experience working with unstructured data for either natural language processing, computer vision or speech applications

• At least 2 years of experience fine-tuning and deploying transformer based models using deep learning libraries and tools such as Pytorch and HuggingFace

• At least 3 years of experience with object oriented Python via experiences in data science and software engineering

New York City (Hybrid On-Site): $165,100 - $188,500 for Princ Associate, Data Science

This role is also eligible to earn performance based incentive compensation, which may include cash bonus(es) and/or long term incentives (LTI). Incentives could be discretionary or non discretionary depending on the plan.

. Eligibility varies based on full or part-time status, exempt or non-exempt status, and management level.

If you have visited our website in search of information on employment opportunities or to apply for a position, and you require an accommodation, please contact Capital One Recruiting at 1-800-304-9102 or via email at . All information you provide will be kept confidential and will be used only to the extent required to provide needed reasonable accommodations.