Senior Machine Learning Engineer

Description

The company is an early-stage, fast-growing business in the Engineering and Product department. It manages over $20 billion in risk and is still early in working out where AI and ML models can deliver the most value.

You will be the second engineer on the dedicated ML team and help shape architecture, tooling and ML strategy.

You will mine large datasets for trends and insights, finding where refined data or internal ML and AI models can improve product results and operations.

The company has no dedicated Data Engineering team, so you will own the data layer for your models. You will often be the first person to work with a raw source, taking it from unexplored data through feature engineering into a production model.

Your work covers public data, proprietary data and large volumes of image data.

Expect roughly 30% of your time on lightweight data pipelines and new dbt tables (not heavy ETL) and roughly 70% on feature engineering, modeling, deployment and monitoring.

You will work with a high degree of autonomy as a trusted technical owner of business problems across the organization. Because the company is still growing, you will wear many hats and help shape product decisions.

Responsibilities

• Own the full ML lifecycle: design, build, deploy and evolve datasets and models, prioritizing scalability, quality and maintainability.
• Build lightweight data pipelines and dbt tables that turn raw data into model-ready inputs, without owning heavy ETL infrastructure.
• Lead the exploration and rollout of new ML applications in the product that predict risk for each insured and across the whole portfolio.
• Work on focus areas such as property-level risk, cost assessment, and aerial image analysis or other modeling that produces attributes for other models.
• Write clean, maintainable code in the existing stack, using AI tools where helpful. The stack is an event-driven architecture on Kafka, AWS (EKS), Python, Django/FastAPI and Postgres, with CI/CD through GitHub Actions. Set a high bar for engineering quality and architecture.
• Work with Engineering, Product, Operations and business teams to design reliable cross-system solutions and confirm your models solve real problems.
• Track model quality, bias and performance with clear metrics that show business value, holding to strong scientific rigor in evaluation.

Requirements

• 5+ years applying machine learning methods to production problems.
• Ability to work in a complex codebase without much ramp-up time.
• Comfortable starting from a raw data source nobody has used before, building a lightweight pipeline or dbt table, then carrying it through feature engineering, modeling and deployment.
• A business mindset: you care about the product side of data, apply modeling and evaluation to real problems, and have clear views on where the data points and how to maximize impact.
• Pragmatic: you balance speed and quality, make sound trade-offs, use off-the-shelf and foundation models, and know when a custom build is justified.
• Curious about how the business works. You spot flawed analyses, suggest ideas nobody asked for, and do not need fully specified specs handed to you.
• Actuarial experience, or experience applying models to risk evaluation and aggregation problems.
• Experience in computer vision and image analysis.
• Experience with dbt or similar modern data transformation tools.
• Plus: past experience as a team lead or owning end-to-end deployment.

Conditions

Full-time position.

Based in the Austin, TX office, 4 days a week on site, near Mopac and W. Anderson Lane.

Benefits

• Pay at the top of the Austin market, plus equity.
• 3 weeks of PTO and 6 federal holidays.
• Medical, dental, vision, life and disability insurance, plus HSA and FSA.
• 401(k) retirement plan.
• Relocation assistance provided.
• Free snacks, team lunches and a collaborative office culture.