Case Study: Bigabid improves production model monitoring and explainability with Fiddler

Q&A with Bigabid CTO Monitoring Thousands of Models in Production

Bigabid, a data company that uses machine learning to power mobile app user acquisition and retargeting, needed a better way to oversee thousands of production models. Before adopting Fiddler, the team relied on an in-house spreadsheet-based process that was difficult to maintain, required daily refreshes, and didn’t provide the visibility needed to track model drift, feature issues, and prediction changes in production.

With Fiddler’s model monitoring and explainability capabilities, Bigabid was able to monitor drift across features and predictions, quickly identify unexpected production changes, and better understand why model outputs shifted. The company reported faster deployment, less manual maintenance, and greater trust in alerts, with issues that previously could take weeks or months to uncover now being detected immediately—helping protect revenue and profit.

Bigabid

Amit Attias

Co-Founder and Chief Technology Officer