MLOps Engineer

Company Description

BLEND360 is adding to our team of passionate people who are revolutionizing the world of Data Science. We are an award-winning, new breed Data ScienceData EngineeringBI Solutions Company focused on powering exceptional results to our Fortune 500 clients. We are a growing company – born at the intersection of advanced analytics, data and technology. 

Our new Director of Data Engineering will be in the forefront of driving major projects within BLEND360 and working with our data engineering team to develop innovative data driven solutions that integrate distributed sources of data, perform large scale learning and reasoning, and integrate results.

Job Description

• Maintain and extend existing data science pipelines in AWS, with an emphasis on infrastructure as code (cloudformation)
• For the purposes of this engagement, extensions will be minimal and limited to those required
• to support the four identified workstreams.
• Maintain and create documentation on infrastructure usage and design (confluence, github
• wikis, diagrams)
• Serve as the internal infrastructure expert, providing guidance to data scientists deploying
• models into the pipelines
• Research new optimization opportunities based on the needs of specific data science products
• Work independently and collaboratively with data scientists to implement optimizations and
• improvements to specific projects deploying or being re-platformed within the infrastructure

Qualifications

• Minimum of 5 years DevOps experience in AWS Cloud including managing ML Pipelines.
• Built and Executed at least 2 MLOps projects in AWS cloud using Sagemaker or other services.
• Experience building cloud infrastructure as code
• Expertise in MLOps best practices
• Foundational understanding of data science and data science best practice
• Experience AWS services (sagemaker, ECR, S3, lambda, step functions) is a must
• Should able write CloudFormation scripts for dev/test/prod environments
• Knowledge in Python
• Should be able to build Docker images independently
• AWS CodeCommit or Github (including github actions) experience is a must