Remote ML Platform Engineer: Feature Store & Deployment

Aimlroles; United States; Remote; USD 180,000 - 230,000; Full time

Benefits offered by this job

Equity

Global onsites

Health plans

Leadership exposure

WFH stipend

Job summary

Tilt is a remote-first fintech company redefining credit through machine learning and scalable platforms. You will own the layer between data science and infrastructure, focusing half the time on a managed feature store and half on training and deployment platforms that empower data scientists to ship models with minimal engineering input.

The role emphasizes building self-service tools, standardizing training pipelines, and enabling rapid model iteration across multiple markets.

Qualifications

• Minimum 4+ years building ML platforms or similar, supporting model deployment at scale.
• Experience with feature stores and model deployment platforms (Databricks Feature Store, Feast, Tecton or equivalent).
• Strong CI/CD for ML and infrastructure as code (Terraform, GitHub Actions, Azure Pipelines).
• Cloud expertise in AWS or Azure for ML deployment, plus Kubernetes and batch/real-time serving.

Responsibilities

• Run and improve the managed feature store and its patterns for defining, backfilling, and serving features.
• Treat the feature store as the starting point for model training, not just deployment.
• Build automated training and deployment platforms used by data scientists across teams.
• Create self-service tooling to automate common ML deployment steps for data scientists.

Skills

ML platform experience

Feature store

Model deployment

CI/CD for ML

Cloud AWS/Azure

Kubernetes

Python/Scala

Data science collaboration

Education

Bachelor's in CS/Math

Tools

Databricks

Feature Store

Terraform

GitHub Actions

Azure Pipelines

Job description

Tilt is a remote-first fintech company redefining credit through machine learning and scalable platforms. You will own the layer between data science and infrastructure, focusing half the time on a managed feature store and half on training and deployment platforms that empower data scientists to ship models with minimal engineering input.

The role emphasizes building self-service tools, standardizing training pipelines, and enabling rapid model iteration across multiple markets.