It all started in 2016 with an app called Crushh, which analyzed text messages to determine how people felt about each other, usually in a romantic context. To do this well, we had to collect users’ full history of text messages (with their consent) across all of their contacts, so we could establish the user’s baseline texting behavior. This allowed us to analyze changes in communication patterns over time. Our app was then able to show users how the balance of a relationship evolved. Shortly after we launched Crushh in the Google Play Store it became the number 1 dating app in US and was featured in hundreds of articles, including 3 of the world’s top 10 publications.

However, we knew the data we collected could have applications outside of romance, even beyond relationships in general. In 2017, we started building our next app, a messaging app called Mei, which has a real-time relationship AI assistant. In addition, we knew that texting on Androids were out of date and decided to build our own implementation of data messaging. If messaging another user, it would recognize that and send via data messaging, but if not, fall back to SMS. We also made it end-to-end encrypted, just like iMessage. In early 2019 we adapted our model to the RCS standard, so if a user’s contact also use Messages Improved (or an RCS 2.0 enabled app) and connected to the internet, messages will be RCS (Rich Communication Services). If not, it will fall back to SMS/MMS. As others open up their implementation, we will make our app compatible.