**3.9 "How to"scenario**

To illustrate the walkthrough, we utilise our prototype software library and further implement a demo application built on top to illustrate plausibility and practical usability of our proposed framework. **Figures 3**–**6** are screenshots taken from the demo application.

#### *Group-Assign: Type Theoretic Framework for Human AI Orchestration DOI: http://dx.doi.org/10.5772/intechopen.96739*

To begin, let us suppose the scenario where we are planning for a fast food restaurant operations. Serving meals to our customers would be core to our business. In this context, "Serve Meal "would therefore be an overall intent from a management perspective. Another person on the team might look at this and suggest that we offer nuggets. Another then contributes that selling burgers will be a great idea too. Now, we have three intents altogether (**Figure 3**): "Serve Meal" and two constituent intents, "Serve Nugget" and "Serve Burger".

Subsequently, another team member points out that a meal would only be complete with a drink and further contributes "Serve Drink" (**Figure 4**). Here, we see that the framework is flexible to handle contributions of intents from different parties in a distributed manner. We could go on and add more intents, but this will suffice in this walkthrough for now.

At this point, we start to have a semblance of a plan on our food menu strategy and at work, this is what is often referred to as a "high level" view. However, it clearly still lacks further granularity such as:


We decide then that what we serve will depend on our inventory except we will not serve beef nuggets because it is not a norm. Looking through our inventory, we have chicken and beef in our raw meat inventory and coca-cola in our drinks. So, we will make beef burgers, chicken nuggets and coca-cola drinks. In doing this, we have effectively created groups of data based on some criteria and associated them with the corresponding implementation and intent (**Figure 5**).

Finally, we decide to offer 2 types of meal: Beef burger with coca-cola and chicken nuggets with coca-cola. This leads us to having a complete plan (**Figure 6**): We know what we want to do (Intent), we know how to do it (Implementation) and we have the ingredients (Data).

It is clear that different levels of details (i.e. intents, implementations) are often being handled by different workers horizontally as well as vertically within a company hierarchy. Any work operations of some scale will necessitate this division of labour. This is where disconnects will happen because there are no guarantees of higher level details connecting with more granular details, especially more so when the big picture is contributed by many different workers.

#### **Figure 3.**

*Constructing intents in the context of menu planning for a fast food restaurant.*

#### **Figure 4.**

*Adding an intent "serve drink".*

### **Figure 5.**

*Associating intent-data-implementation. Intents are represented as red rectangles, data is represented as blue diamonds and implementations are represented as green ovals.*

## **Figure 6.**

*Completed plan for offering fast food meals.*
