About this role
We're hiring an AI Engineer who can build a fraud model and also build the production system that runs it. Intuit's Consumer and Lending Risk team protects the lending products small businesses and consumers rely on: QuickBooks Capital loans and lines of credit, Credit Karma Line of Credit, and File Now Pay Later.
You'll own a lending product line end to end. You decide what to model and build the streaming and batch features it needs. You train and calibrate the model, deploy it to real-time serving, connect it to the rules engine where policy decisions are made, and monitor it in production.
The role combines AI science and ML engineering. You'll work daily with fraud policy, product, analytics and platform engineering. Your models make real-time decisions on real money.
Responsibilities
Own the production path, end to end
- Data: define and instrument the data taps and event-bus topics your models need, with data contracts and named schema owners.
- Features: build streaming aggregates, request-time features and batch features, with point-in-time-correct replay for training data.
- Training: build reproducible training pipelines with scheduled retraining and recalibration.
- Serving: deploy models to real-time and batch serving with online/offline parity, latency budgets and safe fallbacks.
- Decisioning: integrate scores into the rules engine with policy partners, so each decision records whether a rule, a model or a fallback made it.
- Monitoring: own dashboards, drift and data-quality alerts, decision logs, and on-call for your model's decision path.
Do the science
- Frame fraud problems with policy: patterns, label definitions, label maturity and the cost of each action.
- Build calibrated models (gradient-boosted trees, deep and sequence models) whose probabilities a decision can price.
- Design shadow-mode, champion/challenger and holdout experiments, and read the results with Decision Science.
- Handle what makes fraud data hard: delayed and partial labels, unseen outcomes for declined applications, and fast-changing attacks.
- Document models for model risk management, fairness review and bank-partner validation.
Raise the engineering bar
- Write production-quality code with tests, code review, CI/CD and versioned artifacts.
- Turn what you build once into reusable patterns other AI scientists can self-serve.
Qualifications
- MS or PhD in Computer Science, Statistics, Applied Math, Operations Research or a related field, or a BS with equivalent experience.
- 2+ years of industry experience building ML models, including at least one you deployed and operated in production.
- Strong Python and SQL, plus Spark or a similar distributed framework.
- Supervised learning on tabular data: gradient-boosted trees, logistic regression, evaluation and calibration.
- Experience building data or feature pipelines on event streams (for example Kafka) or a workflow orchestrator (for example Airflow).
- Software engineering fundamentals: Git, testing, code review, CI/CD and containers.
- Clear communication with non-technical partners such as policy and product.
Intuit provides a competitive compensation package with a strong pay for performance rewards approach. This position may be eligible for a cash bonus, equity rewards and benefits, in accordance with our applicable plans and programs (see more about our compensation and benefits at Intuit®: Careers | Benefits ). Pay offered is based on factors such as job-related knowledge, skills, experience, and work location. To drive ongoing fair pay for employees, Intuit conducts regular comparisons across categories of ethnicity and gender.
The expected base pay range for this position is:
Mountain View $158,500 - $214,500