Harmony and integration in technology!
Peki chatbots run on the MLNET algorithm and involve language recognition, verification and correction algorithms. It makes a guess with x% probability when it is suitable for which of the classes it is trained, and directs the incoming sentence to the relevant class, if any, or to the “default” flow. Classes are 100% created by the Peki Bot administrator.
You can define the response of the artificial intelligence as a static message, or you can direct it to the streams you will create using rich media plug-ins. The help desk on pek.io provides information on the capabilities of all rich media kinds.
Artificial intelligence can parse the words in an incoming phrase that you want to understand, so it can skip a question if it also includes information you want to learn about the flow. The parsing process helps artificial intelligence correctly grasp the specific request of the client and control the flow.
Bot History function. You can historically list all the bot correspondences on a user basis.