Happy Sunday and welcome to InsideAI! I'm Rob May, Founder and CTO at Dianthus, and active AI angel investor (send deals!) This week I want to write about an emerging business model I'm hearing about in the AI space.
There is a concept in machine learning called Federated Learning. The basic idea behind federated learning is that you don't need centralized training data but rather, you can use data sets from different places, without mingling them together, and train a model across all of the data. This is important because in a B2B SaaS world, companies don't want their data co-mingled with the data from other companies.
One of the problems in launching AI/ML startups is getting data sets. To get around that, here is a business model I've heard several startups pursuing, although, I've only seen one company so far do this successfully at scale.
1. Build a version of a popular tool and make it free.
2. Use federated learning on the data you get from the free customers to build a model that has some commercial value.
3. Sell the model (maybe a more predictive version of whatever the free version does) as a paid version of the service.
I think an appropriate name for this business model is "federated freemium." And I think it can work really well in the AI space. If you are pursuing a federated freemium business model, I'd love to hear from you.
Thanks for reading.
@robmay
Comments
Post a Comment