Full-Time Lead Data Scientist

Job Description:

1. Bayesian Modeling & Marketing Analytics:

• Proficiency in Bayesian inference and modeling, including prior knowledge incorporation and validation techniques.
• Ability to handle seasonality, long-term trends, and multi-touchpoint attribution in Marketing Mix Models (MMMs).
• Addressing collinearity in marketing channels.

1. Big Data & ETL:

• Hands-on experience with tools like Databricks, Spark, and BigQuery.
• Expertise in designing and optimizing data pipelines for marketing data.
• Knowledge of real-time streaming data and partitioning strategies.

1. General Machine Learning Fundamentals:

• Understanding the bias-variance tradeoff, class imbalance, feature selection, and cross-validation.
• Outlier detection and management techniques.

1. Python (sklearn) and Classical ML Techniques:

• Experience with sklearn, including estimators, transformers, pipelines, and hyperparameter tuning.
• Knowledge of ensemble methods and handling categorical variables.

1. LLMs & Generative AI:

• Experience with large language models (LLMs), including prompt engineering and retrieval-augmented generation (RAG).
• Use of LLMs for coding, research, and productivity enhancement.

1. Neural Networks & PyTorch:

• Designing architectures for time-series forecasting.
• Implementation of custom loss functions, attention mechanisms, and optimization algorithms.
• Handling overfitting, learning rate scheduling, and batch normalization.

1. Optimization Techniques:

• Knowledge of optimization methods (e.g., gradient descent, convex vs. non-convex optimization).
• Handling constraints, regularization, and competing objectives.

1. Model Interpretability & Explainability:

• Use of interpretability tools (e.g., SHAP, LIME) and counterfactual analysis.
• Communicating feature importance and debugging complex models.

1. Solution Architecture:

• Designing end-to-end marketing systems covering data ingestion, processing, modeling, and visualization.
• Scalability, versioning, compliance, internationalization, and A/B testing within marketing frameworks.

1. Practical Problem-Solving:

• Debugging model issues, missing data handling, and performance optimization.
• Design and implementation of recommendation systems, anomaly detection, and real-time optimization models.

Preferred Experience

• Hands-on work with marketing teams, developing marketing mix models, or related analytical roles.
• Familiarity with tools and frameworks used in marketing data processing and modeling.