Machine learning (ML) is becoming more commonplace, and can improve the workflow of many different kinds of professionals, whether they be photographers who use ML to stylize photos or dermatologists who use ML to diagnose skin cancer. However, for non-ML specialists, finding and using these models can be quite difficult, as models are scattered across GitHub, personal pages, and hosting services. They are usually hidden behind REST APIs, so that people without a software engineering background (the usual intended users) cannot access them. We have created NumberSeven to make it easier for anyone to discover and use ML models without any coding required.

What it does

NumberSeven is a search engine for machine learning models, but it's more than that. It takes the REST APIs that are provided for ML models and puts a visual front-end in front of the APIs. That makes the models usable for people without software engineering backgrounds. Finally, because many ML models are most usable on the go, we have created a companion mobile app that allows people to access and use the mobile app from their phone.

What's next for

We are adding more ML models to NumberSeven, particularly those in the medical domains. Dermatologists who want to use ML models from their phones are our first target customers, so we will be marketing and selling to doctors and hospitals in the near future.

Share this project: