Email Text Classification: Building An End-To-End Data Product (Return Path) Data Platforms 2018

What You’ll Learn:

1. Introduction to Return Path:
   • Gain insights into how Return Path leverages vast email data to optimize email marketing programs, ensuring content reaches the right audience at the right time.
2. The Full-Stack Data Scientist:
   • Understand the multifaceted role of data scientists who blend expertise in mathematics, statistics, software engineering, and domain knowledge to innovate solutions for critical client challenges.
3. Data Product Lifecycle:
   • Explore the critical phases of the data product lifecycle, emphasizing the importance of discovery, operationalizing, and the iterative nature of the process.
4. Case Study – Email Text Classification:
   • Follow the journey of developing an email text classification project from concept to deployment, highlighting the use of Qubole for efficient data processing and model development.
5. Phase 1 Highlights and Transition to Phase 2:
   • Discover the phased approach that facilitated early wins and strategic planning for expanding the project scope based on solid data foundations and business buy-in.
6. Discovery and Data Preparation:
   • Delve into the discovery phase, where the team collaborates across functions to define project goals and prepare data sets for model training.
7. Modeling and Tuning Tools:
   • Learn about the selection and tuning of machine learning models using tools like Scikit-learn, Spark, and Qubole’s scalable environment for model optimization.
8. Deployment Strategies:
   • Examine strategies for deploying the model in a production environment, leveraging tools like Airflow for managing dependencies and scheduling tasks.
9. Empowering Data Science Teams:
   • Reflect on the process and propose a dream team structure that promotes cross-functional collaboration, with data scientists, product managers, and data engineers working in concert to accelerate innovation.