8 days ago
Base Salary
$221k - $300k/yr
Responsibilities
- Build tools, infrastructure, and workflows that improve the speed and reliability of machine learning development.
- Design and scale a unified embeddings platform for training, serving, and managing representations.
- Streamline feature engineering pipelines and automate data refresh, retraining, evaluation, and drift monitoring.
- Scale training pipelines for larger datasets, more complex architectures, and faster experimentation.
- Improve the end-to-end ML lifecycle from data readiness and feature development through deployment and monitoring.
- Explore algorithms and methodologies and develop supporting tooling for machine learning models.
- Define the roadmap and technical direction for the next-generation ML Platform.
- Collaborate with ML scientists, engineers, Data Engineering, ML Platform, Pricing, and product teams.
- Lead technical initiatives and advise across disciplines in ambiguous, high-impact domains.
Requirements
- 5–7+ years of hands-on experience in applied machine learning with exposure to production-scale modeling.
- Strong theoretical and practical foundation in machine learning and statistics, including model assumptions, bias, uncertainty, tradeoffs, evaluation, and failure modes.
- Deep understanding of model behavior beyond common tools and frameworks and the ability to apply that knowledge to production solutions.
- Experience in high-scale, ML-driven product environments, especially fintech, pricing, or risk modeling.
- Proficiency in Python and core machine learning frameworks such as PyTorch, TensorFlow, Scikit-learn, and XGBoost.
- Experience with end-to-end model development, including data preparation, feature engineering, training, evaluation, and deployment.
- Ability to work autonomously, lead technical direction, and influence strategy across engineering and science teams.
- Experience collaborating with ML scientists, engineers, and product partners.
- Master’s degree or PhD in a quantitative discipline, or equivalent additional professional experience.
- Preferred experience with CUDA/GPU acceleration, feature stores, embedding architecture, synthetic data generation, experimentation frameworks, hyperparameter tuning, and automated model selection.
- Demonstrated ability to improve production model accuracy with measurable business outcomes.
Benefits
- Remote-US position with East/West Coast time-zone requirements.
- Digital-first work model with most work completed remotely and team onsites typically once or twice per quarter for 2–4 consecutive days.
- Competitive base pay, bonus opportunities, and annual equity grants.
- 401(k) or Group Retirement Savings Plan with company matching up to $15,000 annually in the US or CAD equivalent in Canada.
- Employee Stock Purchase Plan for eligible US employees.
- Medical, dental, vision, wellness resources, and Health Savings Account contributions for eligible US plans.
- Life insurance, disability coverage, paid time off, sick leave, company holidays, and paid family and parental leave.
- Family-centered fertility, parenthood, and caregiving benefits plus an Employee Assistance Program.
- Financial wellness resources, annual wellness allowance, and annual productivity allowance.
- Team events, employee resource groups, and onsite office perks including catered lunches and stocked micro-kitchens.
Categories
About Upstart
Upstart builds an AI-powered lending marketplace and cloud software that connect consumers to banks and credit unions for personal, auto, and home-equity credit. Lenders use Upstart’s credit models and applications to automate originations, while Upstart earns platform and referral fees. Founded in 2012 and headquartered in San Mateo, California, Upstart is a public company listed on NASDAQ under the ticker UPST.