🧐 Why Anovos?
Data science teams spend up to 80% of their time on feature engineering and still end up building models with poor resilience. Anovos seeks to address both these issues (lack of modeler productivity and insufficient model resilience) by enabling data scientists to understand the data they work with and to engineer reproducible and robust features. In turn, these serve as the foundation for the training of resilient models that perform reliably and consistently when deployed.
Unlike current feature engineering workflows, which are ad hoc, error-prone, and modeler-driven, Anovos seeks to inject process-driven efficiency into feature creation based on innovations in understanding the stability of data and how data items impact the features they anchor.
With Anovos, ML models become more consistent, more accurate, and deliver results faster. At the same time, the process of building models becomes more projectable, saving time and decreasing cost.
👥 Who's behind Anovos?
Anovos is built by a team of highly talented and experienced data scientists at Mobilewalla with years of experience in applying ML techniques to some of the most extensive consumer data sets available. Frequently, the team found a need to create novel tools to simplify and speed up the feature engineering process to increase efficiency. Through open sourcing these tools, we share our learnings with the community.