Sr. ML Engineer

Company Description

Blend is seeking an experienced Machine Learning Engineer with deep expertise in AWS-based ML pipelines, MLOps best practices, and infrastructure-as-code. This role is focused entirely on pipeline engineering and infrastructure optimization — no model training or research — and will play a critical part in refactoring mature ML systems to support upcoming business initiatives.

The engineer will work closely with cross-functional data science, data engineering, and platform engineering teams to refactor, migrate, and scale production-grade ML pipelines that power recommender systems and lower-priority NLP applications. The ideal candidate will be comfortable with large-scale AWS-native environments, feature store integrations, and high-performance CI/CD workflows for ML.

Job Description

Pipeline Refactoring & Optimization

• Redesign and refactor existing ML pipelines to improve scalability, maintainability, and operational efficiency.
• Migrate pipelines to accommodate new input datasets that will drive updated models.
• Ensure pipelines can handle both batch and streaming workloads.
  
  

Feature Store Integration

• Work with Tecton to manage and serve online/offline features for ML models.
• Migrate legacy feature ingestion and retrieval processes to Tecton.
  
  

AWS Cloud Engineering & Automation

• Develop and manage infrastructure using AWS CloudFormation and other IaC tools.
• Leverage AWS services such as SageMaker, Lambda, ECR, S3, and DynamoDB for ML workflows.
  
  

MLOps & CI/CD

• Implement and maintain deployment pipelines using AWS CodePipeline.
• Ensure seamless integration of ML workflows with SageMaker for training, inference, and monitoring.
• Apply robust testing strategies, code coverage, and quality controls to all ML pipeline code.
  
  

Data Engineering Support

• Integrate ML pipelines with Snowflake, S3, and DynamoDB data sources.
• Optimize data ingestion, transformation, and delivery to production models.
  
  

Technical Debt & Migration Projects

• Identify and remediate technical debt in ML infrastructure.
• Support migration of existing code to align with new feature store and data input requirements.
  
  

Qualifications

Professional Experience:

• 5+ years of hands-on ML engineering experience (7+ preferred).
• Proven success in AWS-based ML pipeline engineering at scale.
  
  

Core Technical Skills

• AWS SageMaker, CloudFormation, Lambda, ECR, S3, DynamoDB.
• Python for pipeline development, automation, and integration.
• Snowflake data integration and optimization.
• CI/CD in AWS CodePipeline for ML workflows.
• MLOps best practices for production-grade pipelines.
  
  

Domain Expertise

• Experience with recommender systems (primary use case).
• Familiarity with NLP applications (secondary focus).
• Strong understanding of batch and streaming ML pipeline architectures.
  
  

Soft Skills

• Ability to work independently on complex refactoring and migration projects.
• Excellent collaboration skills with cross-functional teams.
• Strong problem-solving and documentation capabilities.
  
  

Additional Information

The ML platform supports critical personalization, recommendation capabilities, loyalty programs, and operational optimization. The current ML infrastructure is mature but requires strategic refactoring and migration to handle upcoming product demands.

The Primary Focus For This Role Will Be

• Refactoring recommender system pipelines to incorporate new feature inputs from additional data sources.
• Migrating select pipelines to use Tecton as the centralized feature store, replacing legacy feature engineering paths.
• Ensuring pipelines can operate efficiently in both batch and streaming contexts.
• Maintaining full AWS-native deployments with no hybrid/on-prem dependencies.
  
  

Secondary projects may involve enhancing NLP-related pipelines, optimizing infrastructure automation, and addressing technical debt across existing ML codebases.

This is a critical, high-impact engineering role that will directly shape the company’s ability to deploy faster, more reliable, and more intelligent ML-powered features at scale.