Link or Source to start with Cortex and Machine Learning

We are starting to work with Machine Learning with Cortex where we required to understand the process to use this. Is there any Link or Source to start and Implement with Cortex and Machine Learning? 

  • Sitecore Cortex has not been released. The marketing and sales teams are promoting it, and it's exciting. But to date we are on v9.0 and the best we have is segmented personalisation. No AI / ML ... Maybe next year
  • @Jay Khan is partially correct. Cortex is an umbrella brand which covers all of our machine learning and algorithm capabilities and is already available to you in the existing product in things like Path Analyzer, Personalization, Content Testing...

    However, the pieces everybody have been waiting for in 9.1 are Content Tagging, Personalization Suggestions, and the new Cortex Processing Engine. These are not released yet and therefore there is no documentation available to you for digging deep into it. We will be releasing a lot more content soon about all the new features in 9.1 when the release is made available for download.

    There is a video introduction that you can check out on the website: 

    If you attended Symposium, you can also download some presentations from the speakers in the technical track (Colin te Kempel and Alistair Deneys specifically) to learn more about the Cortex processing engine.

    There are also ways to incorporate ML into your solution without using the Cortex processing engine. I would recommend that if you are new to ML, the first things to figure out are:

    1. What are you trying to do with ML? What business problem are you trying to solve? Doing ML just because is not a recipe for success.
    2. Are you looking at insights on customer behaviour? Or are you looking to automate marketer actions? Predicting outcomes is a very different thing than doing something like Content Tagging, so you need to know who it is you are trying to help.
    3. Is your data ready for ML? You need good, clean data in order to be able to run algorithms.
    4. Are you collecting information about your users that would be helpful in making a decision? You need enough information about behaviour to be able to create models and learn about your users. If you haven't set up xDB yet or learned about Contacts and Interactions, you should start there. You will most likely be trying to do something with your xDB data.

    Hopefully that can get you started until the official documentation is released.