Data Scientist

Key Responsibilities

• Design and implement machine learning models to address business challenges using structured and semi-structured data.
• Create Proof of concepts (POCs) / Minimum Viable Products (MVPs), then guide them through to production deployment and operationalization of projects
• Develop and embed automated processes for predictive model validation, deployment, and implementation
• Work on multiple pillars of AI including cognitive engineering, conversational bots, and data science
• Advanced skills with statistical/programming software (e.g., R, Python) and data querying languages (e.g., SQL, Hadoop/Hive, Scala)
• Good hands-on skills in both feature engineering and hyperparameter optimization
• Experience producing high-quality code, tests, documentation
• Build end-to-end ML pipelines in collaboration with data engineering and product teams.
• Use Google Vertex AI components such as:
• Vertex Pipelines
• AutoML and Custom Training
• Vertex AI Workbench / Notebooks
• Vertex Feature Store
• Optimize model training, deployment, and monitoring workflows using MLOps best practices.
• Translate business requirements into analytical models with measurable outcomes.

Required Skills

• 2-4 years of experience in data science and applied machine learning. Expertise in:
• Python, SQL, R, TensorFlow or PyTorch
• Feature engineering, model tuning, evaluation, and interpretation
• 1 year of experience working with Google Vertex AI or similar GCP ML tools.
• Strong grasp of machine learning concepts: supervised/unsupervised learning, recommendation engines, time series forecasting, etc.
• Exposure to BigQuery, Cloud Storage, or Dataflow for large-scale data manipulation.

Preferred Skills

• Vertex AI Pipelines and Feature Store
• ML lifecycle automation and CI/CD for ML (MLflow, TFX, etc.)
• Docker, Airflow, or Cloud Composer for orchestration
• Google Cloud certifications (e.g., Professional Data Engineer, ML Engineer) are a plus.

Interested candidates can share resume at- ashurustagi@kpmg.com