MLOps Engineer

Job Posting Title

MLOps Engineer

Position Overview:

You Will:

1. Focus on model deployments & performance monitoring, training & inference cost optimizations, tech stack upgrades & migrations.
2. Design, improve, and implement best practices in continuous integration and delivery pipelines to speed the rate at which our machine learning engineers can go from ideation to operationalizing machine learning models in a repeatable manner.
3. Setup new model training pipelines ensuring proper tagging, monitoring, alerting, and logging are in place.
4. Deploy projects into a streamlined environment strategy ensuring they can take advantage of best practices.
5. Identify cost savings in areas where computation and storage utilized for model training and inference can be optimized.
6. Support existing day to day requests through our service desk raised by our platform's customers to quickly provide access, unblock issues, and provide general guidance on best platform practices and patterns.
7. Collaborate with existing machine learning engineers, data scientists, and ML platform engineers from cross-functional teams.

You Bring:

1. Advanced degree (Master’s or above) in Computer Science, specializing in software engineering or platform engineering.
2. 3 years of industry experience working in an MLOps position.
3. Strong programming skills in languages such as Python, Scala, and Java.
4. Hands-on experience in Databricks, mlFlow, Seldon.
5. Excellent problem-solving skills and analytical skills.
6. Deep knowledge of the AWS ecosystem.
7. Expertise in Git, Docker, Kubernetes, Jenkins, Prometheus, and Grafana.

Preferred Qualifications:

1. Experience in supporting as a platform engineer, large-scale online customer-facing ML applications, preferably recommendation systems.
2. Experience working with custom ML platforms, feature store, and monitoring ML models.
3. Familiarity with best practices in DevOps, MLOps.
4. Knowledge of additional CI/CD tooling such as Prow, Bazel, and ArgoCD.
5. Experience in Snowflake, Kubeflow, Tecton.