Challenge

Evaluating more than 5,000 FBA companies, acquiring over 100 top-rated brands, and managing the over 14,000 category-leading products.

Solution

Collect Data from multiple data sources such as Amazon seller central, Walmart, Shopify, direct store sources, and proxy vendor API targets such as Rainforest, Fivetran, Rakuten, etc. 

All efforts start with a Jupiter Notebook in SageMaker, where we do our Ad-Hoc research process, evaluation data, and building models.

There are 3 components that come from the research process: data processing steps, models, and visualizations.

These steps are then being converted into a production-grade code in GitHub Repos, deployed with the CI as Airflow data pipelines or models and visualizations in Plotly Dash.

Result

The team provided M&A, Marketing, SP&A, and Brand Management teams with a reliable foundation for data-driven decision-making, and develop solutions that allow successfully and proactively manage and expand Thrasio’s portfolio of products at scale.