Job Description : Senior Data Scientist (Time Series Forecasting & MLOps)
Location : (Hybrid - Bangalore)
About the Role :
Were searching for an experienced Senior Data Scientist who excels at statistical analysis, feature engineering, and end to end machine learning operations. Your primary mission will be to build and productionize demand forecasting models across thousands of SKUs, while owning the full model lifecyclefrom data discovery through automated re training and performance monitoring.
Key Responsibilities ML Algorithms :
- Design, train, and evaluate supervised & unsupervised models (regression, classification, clustering, uplift).
- Apply automated hyper parameter optimization (Optuna, HyperOpt) and interpretability techniques (SHAP, LIME).
Data Analysis & Feature Engineering :
Perform deep exploratory data analysis (EDA) to uncover patterns & anomalies.Engineer predictive features from structured, semi structured, and unstructured data; manage feature stores (Feast).Ensure data quality through rigorous validation and automated checks.Time Series Forecasting (Demand) :
Build hierarchical, intermittent, and multi seasonal forecasts for thousands of SKUs.Implement traditional (ARIMA, ETS, Prophet) and deep learning (RNN / LSTM, Temporal Fusion Transformer) approaches.Reconcile forecasts across product / category hierarchies; quantify accuracy (MAPE, WAPE) and bias.MLOps & Model Lifecycle :
Establish model tracking & registry (MLflow, SageMaker Model Registry).Develop CI / CD pipelines for automated retraining, validation, and deployment (Airflow, Kubeflow, GitHub Actions).Monitor data & concept drift; trigger re tuning or rollback as needed.Statistical Analysis & Experimentation :
Design and analyze A / B tests, causal inference studies, and Bayesian experiments.Provide statistically grounded insights and recommendations to stakeholders.Collaboration & Leadership :
Translate business objectives into data driven solutions; present findings to exec & non tech audiences.Mentor junior data scientists, review code / notebooks, and champion best practices.Minimum Qualifications :
M.S. in Statistics (preferred) or related field such as Applied Mathematics, Computer Science, Data Science.5+ years building and deploying ML models in production.Expert level proficiency in Python (Pandas, NumPy, SciPy, scikit learn), SQL, and Git.Demonstrated success delivering large scale demand forecasting or time series solutions.Hands on experience with MLOps tools (MLflow, Kubeflow, SageMaker, Airflow) for model tracking and automated retraining.Solid grounding in statistical inference, hypothesis testing, and experimental design.Preferred / Nice to Have :
Experience in supply chain, retail, or manufacturing domains with high granularity SKU data.Familiarity with distributed data frameworks (Spark, Dask) and cloud data warehouses (BigQuery, Snowflake).Knowledge of deep learning libraries (PyTorch, TensorFlow) and probabilistic programming (PyMC, Stan).Strong data visualization skills (Plotly, Dash, Tableau) for storytelling and insight communication(ref : hirist.tech)