Senior Machine Learning Engineer, ALT

• 7+ years of engineering experience, with 5+ years building and shipping production ML/AI models
• Deep proficiency in production-grade Python and SQL, including building custom feature-engineering pipelines (not just off-the-shelf scikit-learn). Think time-decay weighting, leakage-safe k-fold cross-validation, and cascading fallback/imputation logic
• Experience training and validating gradient-boosted or ensemble estimators against strict accuracy/error tolerances, with segment-specific tuning (e.g., by category or asset type)
• Experience leveraging LLMs, foundation models, and AI dev tools for both internal tooling and user-facing product use cases in production
• Experience with MLflow or a comparable tool for experiment tracking and model registry/versioning
• Comfortable owning production model-serving infrastructure on AWS — capacity planning, auto-scaling, and diagnosing memory/timeout failures at scale
• Experience with CI/CD pipelines, orchestrating production workflows, and IaC for provisioning and modifying cloud infrastructure
• Pragmatic and focused on delivering value incrementally rather than pursuing perfection

Desirable

• Optimize our pricing models to significantly reduce infrastructure costs while maintaining and improving their accuracy, especially for high-value assets
• Iterate on our underwriting model to maximize cash advance disbursements while maintaining target risk thresholds and default rates
• Lead the full ML lifecycle from model training and feature generation to production deployment and monitoring
• Collaborate closely with our Expert Pricers to become a domain expert in the trading card market and inform model improvements
• Design and execute experiments and backtesting to discover and validate new features that improve the models’ predictive power and coverage
• Own the models’ AWS infrastructure, writing code for our pricing APIs to ensure the models can serve at scale and with low latency
• What Great Looks Like (6 Months)
• Shipped leaner, more accurate pricing models. You've cut infrastructure cost meaningfully while improving accuracy, especially on high-value assets
• Moved underwriting from good to great. You've iterated on the underwriting model to increase cash advance disbursements without breaching risk thresholds
• Earned trust with Expert Pricers. You're a go-to partner for the pricing team — you understand the domain deeply enough that your model changes reflect real market judgment, not just data
• Hardened the production path. The pricing APIs are faster, more observable, and easier to reason about, with monitoring in place to catch model drift or degradation before it hits customers
• Metrics You’ll Own:
• Northstar Metric: Model-Based Pricing Coverage (% of cards confidently priced by models vs. manually)
• KPIs:
• Pricing Accuracy (% Error)
• Pricing Freshness (End-to-End Model Orchestration Time)
• Underwriting Performance (Advance disbursement rate vs. Target default rate)