Вакансия MLOps Engineer

Skills

• AWS SageMaker
• MLOps
• Python
• CI/CD
• EMR
• AWS Glue
• Athena
• Apache Spark
• LLM
• AWS Bedrock
• Apache Airflow
• AWS CloudFormation
• Terraform
• Salesforce

Описание вакансии

We’re seeking an MLOps Engineer to join AI projects involving machine learning infrastructure, pipeline development, and modern cloud-native tooling. You’ll work with AWS SageMaker, Bedrock, and related services to automate model training, deployment, and monitoring — while optionally supporting data engineering workflows and cloud infrastructure design.
Depending on the project, your tasks will range from building robust ML pipelines to supporting LLM-based analytics agents with real-time data.

Responsibilities:

• Build and maintain MLOps pipelines using Amazon SageMaker (training, evaluation, inference);
• Integrate with CI/CD systems (e.g., CodePipeline, CodeBuild) to deploy models and pipelines;
• Configure monitoring for data/model drift using SageMaker Model Monitor and CloudWatch;
• Collaborate with data scientists and engineers on training, tuning, and deployment workflows;
• Apply ML experiment tracking (e.g., SageMaker Experiments or MLflow);
• Assist with guardrails and quality checks in LLM-integrated applications;
• Document processes, troubleshoot issues, and participate in sprint planning sessions.

Required skills:

• 2-4 years of experience in MLOps, ML engineering, or cloud-based ML deployment;
• Hands-on experience with AWS SageMaker and core MLOps tools;
• Strong Python skills, especially around ML and automation;
• Familiarity with CI/CD practices in cloud environments;
• Understanding of model lifecycle management (training → deployment → monitoring).

Nice to have:

• Experience with data pipeline tools: Amazon EMR, Glue, Athena, Spark;
• Exposure to LLM services (e.g., AWS Bedrock) and vector search solutions;
• Knowledge of Airflow, feature stores, and API-based agent interfaces;
• Infrastructure-as-Code experience (e.g., CloudFormation, Terraform);
• Familiarity with Salesforce or similar data source integrations.