Rize is an intelligent time tracking app that makes its users more productive. It helps you be more in control of your workday and builds better habits by encouraging you to be more focused and more efficient.
Rize automatically tracks and categorizes work activity in real-time to help users to understand how their time is spent, how to reclaim focus time, and how to optimize breaks and meetings. In order to build the best possible time tracker, Rize was looking to implement machine learning to power new features within the product.
Rize is an intelligent time tracking app that makes its users more productive. It helps you be more in control of your workday and builds better habits by encouraging you to be more focused and more efficient.
Rize automatically tracks and categorizes work activity in real-time to help users to understand how their time is spent, how to reclaim focus time, and how to optimize breaks and meetings. In order to build the best possible time tracker, Rize was looking to implement machine learning to power new features within the product.
Rize is the leading personal time tracking and productivity app because of the insights that it generates for its users. By implementing machine learning to rate focus sessions, Rize wanted to improve recommendations to users about their work habits and help them to be even more productive.
Rize’s Postgres database houses millions of records about user focus. Rize wished to improve their rating of user sessions in order to provide even better recommendations to its user base.
The Rize team, while very strong technically, had no knowledge of machine learning. In addition, the timelines and sheer work of building new data pipelines to support such an endeavor seemed daunting. They initially estimated a timeline of at least six months of work, potentially hiring expensive outside ML support to deliver.
Rize quickly saw that connecting their database directly to MindsDB would save them the time and effort of building and maintaining data pipelines and model production architecture for ML.
In addition, using a managed version of MindsDB Cloud would greatly simplify the entire project. Using MindDB Cloud would take all of the infrastructure burden away, meaning that Macgill could focus on delighting users and not messing about with MLOps. The icing on the cake was the knowable pricing component with no hidden risks.
Rize connected MindsDB Cloud directly to the data in their database. From there, they used MindsDB’s AutoML to train candidate models, quickly iterating on a baseline model to one that exceeded their target success criteria. As MindsDB models auto-deploy as DB tables, some simple SQL queries were sufficient to get predictions into the Rize app and start generating value for users.
The overall solution that the Rize team has implemented offers a few key benefits to their business:
“Using MindsDB Cloud to bring machine learning to Rize has been a game-changer. When I started exploring ML, it seemed like I’d need six months of experimentation, model training and building out pipelines to productionize. By using MindsDB Cloud to train and auto-deploy models directly from my database, I saved months of work and many thousands of dollars.”