What do you like best about the product?

Three different crowds use Tensorflow : Researcher , data scientist and developer. Ideally they there were no so such which can collaborate three of them and provide better efficiency. Tensorflow was the solution they created to solve this problem. The library was to scale, it was made to run on multiple CPU's or GPU's and even mobile OS and it has several wrappers. From installation to deployment Tensorflow is the easiest among all the other platform which provide cross cross-platform deployment. Tensorflow provide decent event handling, graph management, image management and best feature is tensorboard.

What do you dislike about the product?

Tensorflow is powerful in terms of performance but with less area to work with. It will restrict you from as tweaking of algorithms in it is quite a complex task. Since the release of Tensorflow 2.0, tensorflow has gain leverage over other framework but its community is still growing and sometime its very hard to find answers when you are stuck in your work.

What problems is the product solving and how is that benefiting you?

For model training and mostly we have used tensorflow as it provide multiple platform deployment.