Drifto automates feature engineering over user events and transaction tables. Drifto aims to provide an automatic point-and-shoot experience: just point Drifto at any number of raw tables, and Drifto automatically combines the raw data into a large number of ML-ready features, aggregated for each ID (e.g. user) over a specified time window (e.g. weekly). Drifto supports workflows such as churn prediction, customer value estimation, anomaly detection, personalization, and demand sensing.
Comments
We've recently launched Drifto.
Drifto automates feature engineering over user events and transaction tables. Drifto aims to provide an automatic point-and-shoot experience: just point Drifto at any number of raw tables, and Drifto automatically combines the raw data into a large number of ML-ready features, aggregated for each ID (e.g. user) over a specified time window (e.g. weekly). Drifto supports workflows such as churn prediction, customer value estimation, anomaly detection, personalization, and demand sensing.
GitHub: https://github.com/drifto-ml/drifto Website: https://www.driftoml.com
We’d love to hear any feedback around how we can better integrate with user data pipelines or thoughts in general.