Azure Machine Learning Workbench
Azure just released a set of updates to Azure Machine Learning. Check out all the details here.
Using Azure Event Grid to Monitor Deployments
Azure just released a cool new service call Azure Event Grid. Azure Event Grid manages all routing of events from any source, to any destination, for any application. It essentially puts events as first class citizens in the ecosystem. Built with a serverless model, it allows you to wire up eventing to perform “shoulder taps” when events happen to trigger downstream processing.
Using Azure Custom Decision Service To Drive Recommendations
I wanted to spice up this blog a little bit with some intelligence. If you notice, on the left (or at the top), there is now a list of recommended articles you might like. This is driven by machine learning algorithms on the backend powered by the Azure Custom Decision Service. This service uses reinforcement learning to personalize the list of links based on your behavior. This means that other users reading the same article may see a different set of recommended articles. The service adapts the list of recommendations to maximize the overall engagement of users.
Using AMQP with Azure Service Bus and Python
Azure Service Bus does support the Advanced Message Queuing Protocol (AMQP) 1.0 protocol. It is a comprehensive messaging protocol and is typically used where reliability and interoperability are key. It provides a wide range of messaging features such as reliable queuing, publish-subscribe, transactions etc.
Creating Context with Microsoft QNAmaker
If you have not checked out Microsoft’s qnamaker then you should. “From FAQ to Bot in minutes” - quite literally. You can sign up, give it an FAQ to ingest, wait a minute or so for it to train and then you will be given a REST endpoint to which you can POST natural language questions. You can quickly incorporate this into a bot using the Azure Bot Framework or even incorporate it into other web applications.