tensorflow-logoSelf-steering cars are a reality, wearable devices can suggest how to improve our sleep, diet, and workouts, we are getting personalized news feeds and search results, and music recommendations that actually make sense. These things are possible through solution based on machine learning. These methods allow a computer to learn from past experience and then make predictions about other quantities. For example, identify an object in a video, translate from one language to another, detect fraud in the bank data, understand that an email you receive is spam and other tasks or improve product offerings to customers.

With Google making TensorFlow open-source, the machine learning technologies can now also be leveraged by companies beyond large corporations. TensorFlow has support for multiple machine learning algorithms and, more generally, can perform any calculations that can be represented by data flow graphs. Machine learning experts praise TensorFlow for rapid development, support across multiple devices, scalability, parallelism and improved debugging experience. TensorFlow is open-source and has extensive documentation. TensorFlow can be used with a simple Python interface and with the right technical background it applications can be developed that integrate in existing business process’.

TensorFlow allows companies allow companies to start experimenting with machine learning in small scale proof-of-concepts with minimal investments. Kentivo has the expertise to help you leverage the potential of TensorFlow to your situation. A typical proof-of-concept would take the approach

  1. Identify a relevant case in the organisation together
  2. Arrange access to the data and iteratively develop a number of machine learning models together and run them against the traditional way of working
  3. Estimate the impact and benefits it might have.

This all can be done within a matter of a few weeks.