pave.dev

Staff Data Scientist

pave.dev United States

Financial Services · 11-50 employees

6 h ago
Remote data-scientist Senior (5-10 yrs) Full-time United States
Create a free account to apply — email only, no card. You can also save this posting or score it against your profile with AI.

About the role

You will own the end-to-end development of credit risk models, taking them from research through A/B testing to production. Additionally, you will build transformer-based foundation models and collaborate with the team to shape AI-native development processes.

What they look for

Data Science Credit Risk Modeling Machine Learning Transformer Architectures PyTorch Gradient-boosted Trees CatBoost XGBoost A/B Testing Feature Engineering Statistical Analysis Data Exploration Model Validation Risk Management Financial Modeling AI Agents

Requirements

Candidates must have strong analytical skills and demonstrated experience taking machine learning models from data analysis to production. Proficiency in deep learning, transformer architectures, and credit risk modeling within the US market is required.

Full description

About Pave

Pave builds cashflow-based credit scores that predict repayment risk more precisely than traditional credit reports, powering a new generation of lending products across consumer and SMB finance. We power nearly 2 billion underwriting events a year to 30+ customers that include SoFi, Happy Money, and Clair.

We’re a small team backed by Better Tomorrow, Bessemer, 8VC, and other top funds and angels from Coinbase, Chime, and CashApp.

The Role

Reporting directly to the Director of Data Science, you’ll build the next generation of credit risk models. At Pave, that means training and/or fine-tuning transformer-based foundation models directly on raw data (e.g., bank transaction, credit reports), in addition to traditional ML models. You’ll work across neo-lending products (cash advance, earned wage access, flexible rent, BNPL) and traditional ones (personal loans, credit cards, small-business financing), taking models from research to live A/B tests with lenders and into production. You’ll also help shape how a small, AI-native team works, using and building AI agents for feature research, model development and evaluation. The models you build will directly move approvals and losses for our customers and shape our product roadmap.

Responsibilities

  • Own credit risk models from customer onboarding through live A/B tests and production.
  • Build transformer-based credit risk and foundation models on raw bank-transaction data, alongside gradient-boosted trees (e.g., CatBoost, XGBoost), and know when each wins.
  • Develop custom credit risk models across neo-lending and traditional products, from cash advance, earned wage access, flexible rent and BNPL to personal loans, credit cards and small-business financing.
  • Make models explainable and bank-ready, including validation and model risk documentation.
  • Communicate findings clearly to customers and teammates through reports, dashboards and presentations.
  • Stay up-to-date with industry best practices and advancements in data science, AI and ML, and credit risk analytics tools, and technologies.
  • Write blog posts about the techniques and tools we’re building.
  • Travel to customers when needed, especially during early deployments.

Requirements

  • Strong analytical and data exploratory skills with the ability to ask the right questions, interpret data and provide actionable insights.
  • Demonstrated experience taking ML models from data analysis to production, including feature engineering and rigorous performance measurement.
  • Hands-on experience with deep learning and transformer architectures (e.g., PyTorch), ideally on sequential or transactional data such as bank transactions or payments.
  • Experience designing and analyzing experiments, such as A/B tests or shadow tests.
  • Strong problem-solving skills and a detail-oriented mindset.
  • Ability to work independently and collaboratively in a remote-first team environment.
  • Ability to clearly communicate findings and insights to both technical and non-technical stakeholders, including external partners and customers.
  • Knowledge and experience building credit risk models, with an understanding of credit risk in the US.
  • Nice to have: experience with model validation and model risk management, fair lending, or training large models on GPUs.

Similar roles