Microsoft Bot Framework
Microsoft Bot Framework

Microsoft Bot Framework provides the developers with a simple and quick way to implement an intelligent chatbot which can converse with the end user, understand its statements and reply with the appropriate result. These bots can be made in Python or C#.

In short, Microsoft Bot Framework personifies the software, making it intelligent via Azure services. Microsoft calls them “Cognitive Services”. These bots can be implemented in any part of a website or be integrated into existing systems to communicate through Email, GroupMe,  Kik, Skype, Slack, Facebook Messenger,  Telegram, Microsoft Teams,  Twilio, Cortana, text/SMS, and Skype for Business.

It took us only 30 minutes to set up a basic Microsoft bot and get it up and running on our development environment. It was much easier than what we expected as we didn’t need to implement the core Artificial Intelligence algorithms.

Cognitive Services use machine learning to enable the bot to see, hear, and understand.

People underestimate bots, if implemented right, a bot can save a lot of time and effort that a human counterpart has to put in, in order to fulfill the end user’s needs.

Here’s a slight insight of what’s in store for developers working on bot framework who will be utilizing the vision api. The cognitive vision API is very interesting. You submit an image to the bot and it will give you a response along with a “Confidence” Value. This value determines how confident the bot algorithm is in recognizing what the image is. A confidence score of 0.99 means it is sure the image contains that object.

If you submit an image of a dog in an indoor environment, it will return you tags such as “Dog:confidence 0.8”, “Indoor: confidence 0.3” and various other objects that it identifies through its algorithms. This means it has identified that the image contains a dog with 80% accuracy.

Based on such confidence values, you can make your own decisions in your application that utilizes the bot’s outcome.


Shriom Manerker

Post Comments

* marked fields are mandatory.