# Introduction to Teneo
Teneo is a Conversational AI workbench that allows you to develop your own conversational application easy, quickly and at scale. 

You can use Teneo to design, develop, connect, maintain and run digital conversations with customers, employees and partners - in the language of your choice, on the channel of your choice.

## Platform Components
The Teneo platform consists of four different components:

- __Teneo Studio__ - the conversational IDE where you develop, test and improve your application. You can also publish your conversational application from within Teneo Studio.

- __Teneo Data__ - your conversational logs are stored in the Teneo Inquire server where they can be accessed for detailed analysis of the conversations and tracking of KPIs and other metrics. 

- __Teneo Language Resources__ - a collection of pre-packaged conversational modules, Teneo Lexical Resources and Teneo Dialogue Resources, that can be used to get your conversational application off to a quick start. 

- __Teneo Interaction Engine__ - the brain behind the run-time natural language processing.

## Platform Architecture
The below image provides a simplified overview of the architecture of the Teneo Platform. Once the conversational application has been built in Teneo Studio, it is published as a web application that can communicate with your end-users on any frontend and in run-time integrate with any backend you may need. Once the conversation ends, the logs are sent to Teneo Inquire for analysis to determine further enhancements of the application.

![Teneo Architecture](//images.ctfassets.net/2vbtnveccz5s/1pj8yGGt3IOOEtMZYLuBKR/be26511ff0a4d6490df64650fffb8a53/Teneo_Architecture.png)

Read more about the platform here [Teneo Features](https://www.teneo.ai/product/features)

Platform overview  - Learning Journey

# From Input to Output

When a user makes an input to a conversation with a conversational application built in Teneo, a lot happens before Teneo identifies the right response that should be returned to the user.

![Input to Output](//images.ctfassets.net/2vbtnveccz5s/7MYYaHcNTtvXs9uHBPqfFm/ae64549c7be687f01bbcbdbf6c3c7d9b/Input_to_Output.png)

In summary, in a project you will have a *solution*. The solution contains your *knowledge base*, where you have *flows* that contains the answer to what an end-user asks. The flows are activated by *triggers* that provide a mapping  between the user inputs on the one hand and the flows in the knowledge base on the other hand. The triggers specify *requirements* that must be met by the user input in order to activate the flow. The first trigger that matches the user input determines which flow will be activated.

### Basic concepts
Below you will find a selection of concepts and features that you will get familiar with and learn more about during the learning journeys:

- __Solution__: a container that holds your project specific knowledge base, the Teneo Engine, your ML model and the assigned TLR
- __Knowledge Base__: a collection of flows, triggers, resources, ordering groups...
- __Flows__: conversational units that contain responses. Can be simple FAQs or complex troubleshooting processes. Flows are activated by triggers.
- __Triggers__: the entry point to a flow. The trigger holds a set of requirements that a user input must meet before the corresponding flow can be activated. A trigger can activate one flow only.
- __Requirements__ : The linguistic, statistical or programmatic requirements on a trigger that the user input must match.
- __Trigger Order__: the ranking of all the triggers which determines the order in which the triggers are evaluated when an input comes in.

From input to output - Learning Journey

# What makes conversations smart?
A truly smart and fluent interaction with a conversational application makes use of a lot of human capabilities, such as:

- __Understanding natural language__. That is, the way a human would take to another human standing in front of him/her. A smart bot understands natural language. It does not force the user to communicate in a specific syntax that does not feel natural. Still, a smart bot also understands the user if they choose to communicate in telegram style or even using only keywords.

- __Having memory__ and knowing what has been said earlier in the conversation, like the user's name for example.

- __Making use of information that has already been provided__ instead of asking for it (again). The bot should only ask for information if the user did not mention it before in the same conversation.

- __Allowing for interruptions__. If a user wants to switch to another topic while in the middle of a complex business process, the bot should allow that and automatically resume the business process once the interruption has been handled.

- __Knowing the context of the conversation__, to be able to recognize short follow up questions that only make sense in a specific context.

- Knowing the limitations of it's understanding and __ask clarifying questions if the user input is not clear__.

- __Being able to engage in conversation about non-business specific content or having some basic conversational skills__, from responding to greetings and compliments to repeating the previous response if requested.


What is in the Platform?

What makes conversations smart?

Next page