About the Role
The Recommendations Data Science Team at Gap Inc. leverages advanced machine learning to power customer experiences across our digital platforms through personalized product recommendations. Our mission is to build scalable, production-ready recommendation systems that support brand strategies and drive engagement and conversion for Gap Inc. and its family of brands. In this high impact role, you will build data pipelines and help to build and deploy deep learning models that power real-time personalization, optimize product ranking, and improve recommendation relevance across our digital platform. You will work closely with GapTech, Product Management, and business partners to ship recommendations optimizations that delight our customers.
What You'll Do
- Develop and deploy automated large-scale ETL pipelines that power state of the art recommendations algorithms
- Collaborate with other data scientists in the team to assist in building AI (ML / Deep Learning) models, iterating on new experimental ideas, and building statistical approaches to evaluate model performance
- Manage unit testing and integration testing frameworks using pytest to ship bug-free code
- Build and maintain dashboards in GCP, Databricks, etc. to monitor the health of the deployed recs systems
- Collaborate with non-technical teams such as Product and brands to translate model performance improvements into potential business impact
Who You Are
- Advanced proficiency using SQL, Python, and Spark for designing and maintaining data pipelines for model training and inference, with a focus on writing production-quality code
- Keen interest to work cross-functionally across MLOps and Engineering to deploy models to production
- Experience with ML and classical predictive techniques such as logistic regression, decision trees, nonlinear regressions, ANN/CNN, boosted trees, SVM, visualization packages, and a track record for creating business impact with these methods
- Prior experience in applying deep learning techniques (using PyTorch and Tensorflow) and contextual bandits highly desirable
- Exposure to Azure, GCP, Databricks, Kubernetes, Cosmos, New Relic, Grafana a bonus