Machine Learning Engineer - Remote Part-Time

4w1 month ago

iScale Solutions

Remote · Part-time · $120,000 – $180,000

About this role

This remote Machine Learning Engineer consultant role focuses on optimizing ML model serving for low-latency inference targeting sub-200ms P95 on EKS. You will advise on and implement AWS-native ML infrastructure including SageMaker endpoints, model registry, A/B testing, and monitoring. The part-time position suits experts in production ML deployment.

Day-to-day tasks include supporting ML-optimized rule weight calibration by training logistic regression or LightGBM on rule-fire indicators from labeled data. Assist with model retraining pipeline automation and drift detection. Contribute to model explainability documentation using SHAP-based attribution for regulatory compliance.

Participate in model governance with version control, audit trails, and threshold configuration per participating institution. Support load testing and performance benchmarking of the ML scoring pipeline. Collaborate on technical proposals and architecture documentation in a flexible remote environment.

Enjoy fully remote work with flexible hours around 15–20 per week. The engagement offers potential extension based on project phase progression. Ideal for certified AWS ML professionals seeking impactful consulting opportunities.

Requirements

AWS Machine Learning Specialty Certification (or AWS Certified Machine Learning Engineer – Associate) — current and valid
3+ years of hands-on experience deploying ML models in production on AWS
Strong Python skills (scikit-learn, LightGBM/XGBoost, pandas)
Experience with containerized ML serving (Docker, Kubernetes/EKS)
Familiarity with model monitoring, drift detection, and retraining pipelines
Experience in fraud detection, AML, or financial risk systems
Knowledge of SHAP or other model explainability frameworks
Experience with SageMaker (endpoints, model registry, pipelines)

Responsibilities

Optimize ML model serving for low-latency inference (target: sub-200ms P95) on EKS
Advise on and implement AWS-native ML infrastructure (SageMaker endpoints, model registry, A/B testing, monitoring)
Support ML-optimized rule weight calibration — training logistic regression / LightGBM on rule-fire indicators to learn optimal rule weights from labeled data
Assist with model retraining pipeline automation and drift detection
Contribute to model explainability documentation (SHAP-based attribution) for regulatory compliance
Participate in model governance: version control, audit trails, threshold configuration per participating institution
Support load testing and performance benchmarking of the ML scoring pipeline
Provide input for the technical proposal and architecture documentation