Data Scientist - PyTorch / TensorFlow

About the role

UPS is hiring a Data Scientist who builds and implements advanced analytics models and solutions that turn large volumes of structured and unstructured data into predictive and prescriptive insights. The role works with a team that researches predictive requirements, using machine learning and data visualization tools in on-premise and cloud environments, and turns data and analytics needs into predictive solutions.

Key responsibilities

• Define key data sources from UPS and external sources to deliver models.
• Develop pipelines for data cleansing, transformation and enrichment from internal and external sources.
• For larger teams, work with data engineering teams to validate and test data and model pipelines from proof of concept.
• Design data structures from exploratory analysis of large data sets to find trends and patterns that meet business needs.
• Set model KPI expectations and handle validation, testing and re-training of existing models.
• Document solutions with project documentation, process flowcharts, logs and commented code, and present findings clearly to stakeholders.
• Apply statistical modeling, machine learning, distributed computing and cloud-based AI best practices toward deployment.

Required profile

• Ability to take a data science problem from start to finish and build the full model product with PyTorch/TensorFlow.
• Strong analytical skills and attention to detail.
• Experience launching moderate-scale advanced analytics projects in production at scale.
• Master's degree in a quantitative field (mathematics, computer science, physics, economics, engineering, statistics, operations research, etc.) or equivalent job experience.

Required skills

• PyTorch and TensorFlow
• R, SQL and Python
• Statistical techniques, machine learning and operations research
• Cloud-AI technologies, Linux/Unix and Windows environments
• Open-source technologies and cloud services