Machine Learning Engineer

Job Specification:

Builds and deploys machine learning models across supervised, unsupervised, and deep learning use cases. Implements feature engineering, hyperparameter tuning, and model evaluation, while maintaining automated training and deployment pipelines using tools like ML flow, Kubeflow, and Airflow. Ensures production reliability through continuous monitoring, performance tracking, and data pipeline optimization.

Key Responsibilities:

ML Model Development

• Develop and deploy ML models (supervised, unsupervised, deep learning).
• Perform feature engineering, model selection, and hyperparameter tuning.
• Implement model evaluation and interpretability techniques (SHAP, LIME).

MLOps & ML Infrastructure

• Build and maintain ML pipelines: training validation deployment.
• Implement experiment tracking and model versioning (MLflow, W&B or Neptune).
• Deploy models using serving infrastructure (TensorFlow Serving, Torch Serve, Triton, BentoML).
• Work with ML orchestration tools (Kubeflow, Airflow, Metaflow).
• Support A/B testing and model deployment strategies.

Model Monitoring & Alerting

• Implement monitoring for model performance: latency, accuracy, drift detection.
• Build dashboards (Grafana, Kibana, CloudWatch) and configure alerts (Slack, PagerDuty).
• Log predictions and actuals for continuous monitoring.

Data Engineering

• Write efficient SQL: joins, aggregations, CTEs, window functions.
• Work with cloud data platforms (Databricks, Snowflake, Big Query, Redshift, Synapse).
• Build ETL/ELT pipelines using AWS Glue, dbt, or Airflow.

Platform Engineering & Observability

• Contribute to ML platforms on Kubernetes/EKS.
• Monitor cloud costs and resource consumption.

Required Skills:

Strong hands-on experience.

ML & Data Science

• EDA and ML algorithms: regression, classification, clustering, ensembles (XGBoost, LightGBM, CatBoost).
• Experience with any deep learning framework: TensorFlow, PyTorch, or Keras.
• Model evaluation and cross-validation techniques.

MLOps

• Hands-on with experiment tracking (any of MLflow, W&B, Neptune).
• Experience with model serving (any of Torch Serve, Triton, BentoML, KServe).
• Familiarity with ML orchestration (any of Kubeflow, Airflow, Prefect, Metaflow).
• Docker and basic Kubernetes knowledge.

Programming & Automation

• Strong Python programming.
• Basic Bash scripting.
• Debugging and troubleshooting skills.

Nice to Have Skills

• Cloud platforms: AWS/GCP/Azure (SageMaker, Vertex AI, Azure ML).
• Feature stores (Feast, Tecton, SageMaker Feature Store).
• Model optimization (quantization, pruning, distillation).
• Distributed training (Horovod, PyTorch DDP, Ray).
• Open-source ML contributions.