Director, Data Science

Walmart Walmart · Retail · Bentonville, AR

Director of Data Science at Walmart responsible for leading a team in the development and deployment of AI/ML solutions for fraud modeling. The role involves managing a team, training and deploying supervised and unsupervised models for real-time fraud detection and anomaly detection, mining large datasets, designing KPI reports, implementing various ML algorithms, conducting data and model drift analysis, and utilizing cloud platforms (Azure/GCP) for scalable deployment. Experience with vector databases and frameworks like LangChain is also required.

What you'd actually do

  1. Collaborates with stakeholders to support user interfaces and promote model usability at scale.
  2. Writes code using appropriate languages (e.g., SQL, Python, Java) based on technical and business needs.
  3. Develops and tests features, creates POCs, deploys software, documents code, and updates progress.
  4. Applies relevant software testing techniques.
  5. Applies best practices in visualization for complex data using tools like Tableau, PowerBI, Python, and R libraries.

Skills

Required

  • Managing a team of data scientists that implement AI/ML solutions in area of Fraud modeling
  • Training and deploying supervised classification models using Python/Spark to detect fraud in real-time
  • Training and deploying unsupervised network & cluster models using Python/Spark to detect fraud anomalies
  • Mining large datasets to discover transaction patterns, examine retail data (orders, sales, returns) and isolate targeted information using traditional & advanced data science techniques
  • Designing and developing KPI reports using SQL & Tableau to monitor real-time performance of models
  • Implementing Machine Learning algorithms including Supervised (Regression/LASSO, SVM, Neural Networks, etc..) and Unsupervised (Association, Clustering, etc.)
  • Conducting data and model drift analysis on production environment to identify inconsistencies & anomalies
  • Utilizing statistical & data management tools & languages (Python, SQL, SAS) to provide insights & data solutions to business problems
  • Working with a team from multiple disciplines, including engineering and product management to build new applications and improve business processes
  • Azure / GCP for scalable deployment containerization (Docker, Kubernetes)
  • Vector databases (Pinecone, Weaviate, FAISS, Milvus)
  • Frameworks (LangChain, AutoGen, CrewAI, Haystack.

Nice to have

  • SQL
  • Python
  • Java
  • Tableau
  • PowerBI
  • R libraries
  • SAS
  • Docker
  • Kubernetes

What the JD emphasized

  • Managing a team of data scientists that implement AI/ML solutions in area of Fraud modeling
  • Training and deploying supervised classification models using Python/Spark to detect fraud in real-time
  • Training and deploying unsupervised network & cluster models using Python/Spark to detect fraud anomalies
  • Mining large datasets to discover transaction patterns, examine retail data (orders, sales, returns) and isolate targeted information using traditional & advanced data science techniques
  • Designing and developing KPI reports using SQL & Tableau to monitor real-time performance of models
  • Implementing Machine Learning algorithms including Supervised (Regression/LASSO, SVM, Neural Networks, etc..) and Unsupervised (Association, Clustering, etc.)
  • Conducting data and model drift analysis on production environment to identify inconsistencies & anomalies
  • Utilizing statistical & data management tools & languages (Python, SQL, SAS) to provide insights & data solutions to business problems
  • Working with a team from multiple disciplines, including engineering and product management to build new applications and improve business processes
  • Azure / GCP for scalable deployment containerization (Docker, Kubernetes)
  • Vector databases (Pinecone, Weaviate, FAISS, Milvus)
  • Frameworks (LangChain, AutoGen, CrewAI, Haystack.

Other signals

  • Deploying models in production
  • Managing a team of data scientists
  • Real-time fraud detection