Deploy the ML models developed by data scientists in the production systems using GCP, BigQuery, and Airflow
Collaborate with data scientists to seamlessly integrate model code (Python/PySpark) into the MLOps workflow.
Resource and code Optimization of the ML model prediction scoring pipelines in GCP ecosystem so that the models run within the defined SLA
Debug the ML model scoring failures for code and resource issues
Orchestrate the model scoring pipelines using Cloud Composer
Deploy Auto-ML solutions in the production systems
Implement model performance monitoring (LIFT, ROC, Accuracy, and so), model stability monitoring, and feature/concept drift monitoring for all the new models
Implement modernization practices such as observability and explainability for enhanced model monitoring and interpretability.
Work closely with Responsible AI teams to ensure ethical AI principles and incorporate fairness, transparency, and accountability.
Utilize Jenkins and GitLab for effective code management and version control.
Leverage Vector DB(Milvus or any other), and LLM services in order to deploy LLM models
Qualifications:
8+ years of proven experience in AI/ML Engineering and minimum of 2 to 3 years of experience in GCP ecosystem is a must
Master's or Bachelor's in Computer Science or in related fields; Concentration in Data Science is good to have
Strong expertise in MLOPs best practices, CICD, and orchestration
Basic to intermediate understanding on the working mechanism of each of the following: Classification (Random Forest, xgboost, catboost), Regression (Linear, Lasso, Ridge), Recommendation systems, dimensionality reduction techniques, deep learning (Neural Networks), and LLMs
Proficient in PySpark and Python programming and GitLab for code management.
Excellent communication skills for effective collaboration.
Candidates holding Certification in Google Cloud and in Data Science are highly encouraged