Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

The privacy concerns with throwing loads of data at a third party api is staggering and perhaps the main driver for doing ML in-house.


This is why I do on premise deployments for my ML stuff. A lot of money is being left on the table with SAAS in this space I think. I'm of the opinion models should come to the data not the other way around.


It's not really an important feature--if it was, people wouldn't have been gushing their details into massive databases over the past decade.

It's nice to see a set of tools that lets even the most simple of business people exploit their customers as effectively as, say, Facebook or Google.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: